article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
d - branes @xcite are indispensable objects in string theories to the understanding of their non - perturbative aspects , including various dualities @xcite . one way to examine string dualities
is to use the d - brane action which is a low energy effective action of massless fields induced on the d - brane @xcite .
for example , taking the worldvolume dual of the d@xmath1-brane action , we obtain a nambu - goto string action with a tension rotated by @xmath2 duality transformation which is conjectured to exist in type iib superstring theory @xcite .
similar analyses have been performed for other cases , such as d@xmath3-brane ( @xmath4 ) actions or super-@xmath3-brane actions @xcite . as another argument on string dualities using d - brane action
we cite ref .
@xcite , which shows that the bps mass spectra given as a volume factor of the d - brane action describing compactified d - branes are consistent with u - duality .
the d - brane actions used in the above analyses are the ones given by the lowest order calculation in the string @xmath0 model approach @xcite , or in the open string partition function approach @xcite , in addition to the ordinary scattering amplitude approach @xcite .
higher order calculations of the d - brane actions have been known through the latter approaches @xcite , in the trivial background of the massless sector of the closed string excitations ( flat metric and no kalb - ramond two - form field ) .
this adds to the d - brane action higher derivative ( @xmath5 ) corrections , which , however , may ruin the dualization procedure of refs .
terms ) are calculated and its consistency with string dualities is checked @xcite . ]
therefore , a systematic framework which gives higher derivative corrections to the d - brane action is highly desired . since d - branes are defined as hypersurfaces on which the boundaries of the string worldsheets can sit , the boundary states @xcite describe the dynamics of the d - branes well . a part of the d - brane action ( more precisely , couplings to the fields of the closed string excitations ) is encoded in the boundary state @xcite . however , this boundary state is defined as an eigen state of the boundary conditions for the oscillating modes of the attached string , the explicit form of the boundary state has been given only for simple backgrounds which give linear boundary conditions .
one of those configurations is the constant field strength on the d - brane .
therefore , boundary states including general configuration of the boundary gauge field give us a method to calculate the higher derivative corrections to the d - brane action . in ref .
@xcite , a d - brane sector was introduced to the system of covariant closed bosonic string field theory ( sft ) @xcite .
the d - brane action corresponds to the source term in this sft , and the boundary state is a constituent of it .
the dynamical degree of freedom of the boundary state adopted there was only the constant field strength . in this scheme of sft ,
a @xmath0 model gauge transformation was realized is such a way that it yields a constant shift of the gauge field strength in the boundary state . in this paper
, we present a generalized boundary state , which is constructed by a gauge transformation in the sft .
this boundary state incorporates the non - constant modes of @xmath6 and gives the derivative corrections to the d - brane actions .
our new boundary state is defined for an arbitrary gauge field configuration @xmath7 , and we carry out the explicit calculations up to the second order in the derivatives on @xmath6 .
this boundary state includes divergences , and the requirement for these divergences to vanish gives a constraint on the configuration of the gauge field @xmath7 .
this constraint is shown to be identical with the vanishing of the @xmath8 function in a string @xmath0 model , hence we regard this constraint as an equation of motion for the gauge field .
after this regularization procedure for the boundary state , we obtain the higher derivative corrections to the d - brane action , by extracting relevant parts in the boundary state .
it is expected that this generalized boundary state serves as an useful tool for the investigation of the dualities or other dynamics in string theory .
the organization of the rest of this paper is as follows . in the next section , after summarizing the properties of the boundary states adopted in ref .
@xcite , this boundary state is generalized to the one which has non - constant @xmath6 boundary coupling . in sec .
[ sec:3 ] , we study the divergence immanent in the generalized boundary state , and see the coincidence with a string @xmath0 model . in sec . [
sec : correction ] , using the new boundary state proposed in sec .
[ sec : extending ] , we calculate the correction to the d - brane action . the final section is devoted to our conclusion and discussions .
we include an appendix where the calculation of the formula appearing in this paper is given .
as mentioned in the introduction , a certain sector of the boundary state is related to the low energy effective action of d - branes @xcite . the boundary state adopted in ref .
@xcite and in most of the literatures has a single free parameter , that is ,
_ constant _ field strength of the gauge field on the d - brane .
therefore the obtained effective action on the d - brane contains only the constant field strength . in this section ,
we generalize the boundary state so that it includes _ non - constant _ modes of the field strength , in order to calculate the corrections to the d - brane action .
these corrections correspond to the derivative corrections , since we expand the field strength by derivatives .
first let us summarize the relevant properties of the boundary state with the constant field strength .
the boundary state @xmath9 is defined as an eigen state of the boundary conditions for open bosonic strings @xcite : @xmath10 the brst invariance of the boundary state is a consequence of eqs.([x_b=0 ] ) ( [ pi_c_b ] ) .
the oscillator representation of @xmath11 reads @xmath12 where @xmath13 the orthogonal matrix @xmath14 is defined as @xmath15 .
although the front factor @xmath16 is arbitrary , this can be fixed by various ways @xcite . here
we shall follow ref.@xcite , where the gauge invariance principle for the d - brane ( source ) term in the closed sft action has fixed the front factor .
only when the front factor is given by @xmath17 the boundary state is subject to the following sft gauge transformation : @xmath18 where the @xmath19 product denotes three string interaction vertex .
a string field @xmath20 is a gauge transformation parameter @xmath21 with @xmath22 and we have defined the gauge transformation for the variable @xmath23 as @xmath24 for verifying the gauge transformation ( [ b*f ] ) , we have used a relation on the @xmath19 product @xmath25 for the above @xmath20 ( [ lambda ] ) .
the gauge transformation for the closed string field is given by @xmath26 therefore with the transformation parameter ( [ lambda ] ) , the inhomogeneous part of the gauge transformation ( [ transp ] ) generates a shift of a component kalb - ramond gauge field @xmath27 in @xmath28 by @xmath29 .
( our notation of sft is given in ref .
this fact is expected from the gauge symmetry of the string @xmath0 model , which is defined as @xmath30 and @xmath31@xcite .
note that in the string @xmath0 model approach , this gauge transformation @xmath32 generates all the modes of the gauge field @xmath7 on the boundary .
therefore , in order to generalize the boundary state to the one incorporating the functional degree of freedom @xmath7 ( or @xmath6 ) , it is natural to demand the following relation for the new boundary state @xmath33 : @xmath34 in this relation the parameter string field @xmath20 is defined by ( [ lambda ] ) for an _ arbitrary _ function @xmath35 , and the gauge transformation law concerning @xmath7 is @xmath36 since the gauge transformation ( [ gaugetr ] ) generates an arbitrary deformation of @xmath7 by the parameter @xmath35 , @xmath37 is obtained by `` integrating '' eq .
( [ transnew ] ) . for this purpose
we use the following formula for the star product @xmath38 between the present @xmath20 and an arbitrary string functional @xmath39 ( the derivation is given in app .
[ app : star ] ) : @xmath40 this is a generalization of the identity ( [ precon ] ) . assuming that the ghost part of the new boundary state is given by the same form ( eq . ( [ ghostf ] ) ) as before , the second term on the rhs of eq . ( [ star ] ) acting on the state @xmath41 vanishes due to the ghost boundary condition ( [ pi_c_b ] ) .
then the desired generalized boundary state @xmath41 is obtained by exponentiating the operator on the rhs of ( [ star ] ) : this definition is very natural in the sense of path - integral approach of string @xmath0 model @xcite .
the boundary state is regarded as an initial or final wave functional for the path - integral .
the rotational operator @xmath42 $ ] is a path - integral weight @xmath43 , where @xmath44 is a minimal interaction term with external gauge fields @xmath45 in the string @xmath0 model action . ]
@xmath46\:{\left| b(f\!=\!0)\right\rangle},\end{aligned}\ ] ] with @xmath47 } u[a]\equiv \exp \left({-i\over \pi}\oint\ !
d\sigma \ , { \partial}_\sigma x^\mu a_{\mu}(x ) \right).\end{aligned}\ ] ] note that @xmath42 $ ] and hence @xmath48 has an invariance under the @xmath49 gauge transformation of @xmath7 , @xmath50 .
the new boundary state ( [ bnew ] ) is reduced to the previous one ( [ bou ] ) when the dynamical variable is restricted to the constant field strength @xmath51 since the neumann part of the previous boundary state ( [ bou ] ) is written in a form @xmath52 our @xmath41 satisfies various desired properties .
one of them is the fact that it obeys the following generalized _ nonlinear _ boundary condition @xmath53 { { \left| { \cal b}\right\rangle}}= 0,\end{aligned}\ ] ] where the field strength @xmath54 is defined as usual : @xmath55 this eigen equation ( [ eigen ] ) is easily proved from a property of the unitary operator @xmath42 $ ] ( [ u[a ] ] ) : @xmath56\pi p_\mu u[a]^{-1 } = \pi p_\mu + f_{\mu\nu}(x ) { \partial}_\sigma x^\nu\end{aligned}\ ] ] and @xmath57 . the new boundary state @xmath41 satisfies a generalized boundary condition ( [ eigen ] ) which is in the same form required in the scheme of the string @xmath0 model @xcite . thus also from this point of view it is possible to assert that this state @xmath41 really describes a boundary of a string worldsheet coupled to a massless gauge field @xmath7 .
one of the other properties of @xmath41 is the brst invariance which is important to understand loop - corrected equations of motion for closed string theory @xcite where we consider only the linearized part .
the equation of motion follows as @xmath58 hence multiplying @xmath59 on this expression we obtain a constraint @xmath60 .
this constraint of the brst invariance of the boundary state is also deduced from the gauge invariance of the whole sft action@xcite . ] .
the brst charge can be written in the form ( see appendix a in the ref .
@xcite for its complete expression ) @xmath61 -c\ ; p_m{\partial}_\sigma x^m \biggr\ } \ ; + \ ; q_{\rm b}^{({\rm ghost})},\end{aligned}\ ] ] where the term @xmath62 consists only of the ghost coordinates .
noting that the ghost part of the new boundary state @xmath63 is the same as the previous one @xmath64 , we have @xmath65 then the commutativity @xmath66=0\end{aligned}\ ] ] ensures the brst invariance @xmath67 although this boundary state @xmath41 is a solution of the equation ( [ brst ] ) for arbitrary configuration of the gauge field , this solution is formal , in a sense that the boundary state contains divergence and thus is not well - defined unless regularized . in the next section we study on this point in detail .
the new boundary state ( [ bnew ] ) for a general ( polynomial ) @xmath7 is made of products of the operators @xmath68 s at the same point @xmath0 and therefore it inevitably contains divergences .
we shall show in this subsection that these divergences can be absorbed into the redefinition of the gauge field as @xmath69 .
we require that the divergent part @xmath70 to vanish , in order to obtain a well - defined boundary state .
eventually , constraint equation is found to be identical with the vanishing of the @xmath8 function of the string @xmath0 model coupling .
thus this requirement of no - divergence coincides with the condition of the conformal invariance in the string @xmath0 model . for studying the divergences in the new boundary state
, we divide @xmath71 into the zero mode and the non - zero mode part @xmath72 , @xmath73 and taylor - expand @xmath74 around the zero mode @xmath75 as @xmath76 we denote the contribution of the second term on the rhs of ( [ aexpand ] ) to @xmath42 $ ] ( [ u[a ] ] ) by @xmath77 and that of the rest terms by @xmath78 , and express @xmath42 $ ] as @xmath79=v u_0.\end{aligned}\ ] ] explicitly , we have @xmath80 and @xmath81 in ( [ v ] ) we have written explicitly only those terms where the total number of derivatives acting on @xmath6 is at most two .
since the operator @xmath77 acting on @xmath82 generates the `` old '' boundary state ( [ bou ] ) with @xmath83-dependent field strength ( cf . , eq .
( [ dividef ] ) ) , @xmath84 the new boundary state ( [ bnew ] ) is expressed as @xmath85 note that the approximation to the first order with respect to the derivative acting on the field strength , where @xmath78 is put to be @xmath1 , does _ not _ reduce @xmath41 to the previous ( brst invariant ) boundary state @xmath9 considered in the ref .
actually , the decomposition of @xmath42 $ ] into @xmath86 does not respect the brst invariance at each order in the expansion of @xmath78 .
next , we shall evaluate the representation ( [ vb ] ) by expressing it only by the creation operators acting upon @xmath87 .
namely , we change all the annihilation operators in @xmath78 into the creation ones with use of the neumann boundary condition satisfied by @xmath87 : @xmath88 where @xmath89 is an orthogonal matrix defined by @xmath90 the calculation is straightforward and we here present only the results ( see app . [
app : div ] for the detailed calculation ) .
the unitary operator @xmath78 reexpressed in terms of creation operators contains two kinds of divergences , @xmath91 adopting the zeta function regularization , @xmath92 , we find that the @xmath93 divergence is absorbed into the redefinition of the field strength @xmath94 : @xmath95 { { \left| b\bigl(f(x)\bigr)\right\rangle}}\end{aligned}\ ] ] where the redefined @xmath96 containing the @xmath93 divergence is given by @xmath97.\end{aligned}\ ] ] here we have kept only those terms written explicitly in eq.([v ] ) .
, since if the corrected action is written by gauge invariant field strength @xmath23 , the number of indices should be even . ] the redefinition ( [ fren ] ) can be expressed in terms of gauge field @xmath98 as @xmath99 for the generalized boundary state to be well - defined ( not divergent ) , the coefficient of the @xmath93 in eq . ( [ aren ] ) should put equal to zero .
therefore we obtain a constraint on the configuration of the gauge field @xmath100 at a glance one recognizes that this @xmath101 is identical with the beta function for the gauge field coupling in the string @xmath0 model @xcite .
hence the constraint ( [ beta ] ) for the generalized boundary state coincides with the conformal invariance condition in the string @xmath0 model .
we shall explain the detailed correspondence between our formalism and the string @xmath0 model approach in the next subsection .
we have seen in the previous subsection that the divergences immanent in the newly defined boundary state coincide with the result in the string @xmath0 model .
this feature is natural in a sense that , in the both formalisms , one evaluates the same weight @xmath102 ( see the footnote [ weight ] ) on the boundary of the string worldsheet .
however , two approaches appear to be clearly different in the following respect : although the @xmath0 model approach gives a loop contribution to the gauge potential itself , the divergences in the boundary state show up in the form of the redefinition of the field strength . in this subsection
we study how the divergences in the boundary state can be understood in the scheme of the string @xmath0 model .
consider a string @xmath0 model action @xcite @xmath103 \right ] , \end{aligned}\ ] ] where the prime denotes the derivative with respect to a worldsheet coordinate @xmath0 .
the boundary of the worldsheet is put at @xmath104 , since we intend to study the correspondence with the boundary state and therefore treat a closed string emitted by the d - brane .
expand the action ( [ ssigma ] ) around a classical configuration @xmath105 as @xmath106 = & & { \hspace*{-5 mm } } \left[\;\ ; { \frac{1}{2}}\int \ ! d^2\sigma\ ; { \partial}_\alpha \bar{x}_m { \partial}^\alpha \bar{x}^m -i \oint \ ! d\sigma\ ; ( \bar{x}^\mu ) ' a_\mu [ \bar{x } ] \;\;\right ] { \nonumber}\\ & & { \hspace*{-5 mm } } + \left [ \;\;\int\ ! d^2\sigma \ ; { \partial}_\alpha \bar{x}_m { \partial}^\alpha \xi^m -i \oint\ ! d\sigma\ ; \xi^\mu(\bar{x}^\nu ) ' f_{\mu\nu}[\bar{x } ] \;\;\right ] { \nonumber}\\ \label{expand } & & { \hspace*{-5 mm } } + \left[\;\ ; { \frac{1}{2}}\int\ ! d^2\sigma \ ; { \partial}_\alpha \xi_m { \partial}^\alpha \xi^m -\frac{i}{2 } \oint\ ! d\sigma\ ; \ ; \xi^\mu(\xi^\nu ) ' f_{\mu\nu}[\bar{x } ] \;\;\right ] { \nonumber}\\ & & { \hspace*{-5 mm } } -i \oint\ !
d\sigma \left({\frac{1}{2}}\xi^\rho \xi^\nu ( \bar{x}^\mu ) ' { \partial}_\nu f_{\rho\mu}[\bar{x } ] \right .
{ \nonumber}\\ & & { \hspace*{10 mm } } + \left . { 1\over 3 } \xi^\nu \xi^\rho ( \xi^\mu ) ' { \partial}_\nu f_{\rho\mu}[\bar{x } ] + { 1\over 8 } \xi^\delta\xi^\rho\xi^\nu(\xi^\mu ) ' { \partial}_\delta{\partial}_\rho f_{\nu\mu}[\bar{x } ] + \cdots \right ) , \ ] ] where the fluctuation around the classical configuration is denoted as @xmath107 and we have adopted euclidean formalism , @xmath108 . in this expansion
, we are working in slowly varying fields , _
i.e. _ derivative expansion . in the above equation ( [ expand ] ) , terms in the second line
vanishes due to the equations of motion for the classical configuration @xmath105 .
the terms in the third line gives a kinetic term ( propagator ) for the scalar field @xmath107 .
this propagator depends on the constant field strength @xmath23 on the boundary . usually in the string @xmath0 model calculation @xcite ,
one evaluates a vacuum graph using the vertex in the fourth line by the contraction @xmath109 } .\end{aligned}\ ] ] this quantity is in proportion to @xmath110 , therefore the divergence of this one - loop vacuum graph is interpreted as the one - loop divergent quantum contribution to the gauge field @xmath111 $ ] in the coupling @xmath112 $ ] which appears in the first line in eq .
( [ expand ] ) . however , when we ignore the first line and only consider the dynamics of the scalar field @xmath113 , it is natural to calculate the renormalization of the propagator , three point vertex , and so on .
let us concentrate on the one - loop correction to the propagator . using the couplings in the fifth line in eq .
( [ expand ] ) , at one - loop level we have two relevant feynman graphs shown in fig .
[ loop1 ] and [ loop2 ] , ( 70,-10)@xmath114 ( 118,40)@xmath115 ( 60,40)@xmath115 which bring out divergences in the sector of the propagator .
though generally the divergence in the two - point function is absorbed into the renormalization factor of the wave function , now according to the philosophy of the string @xmath0 model , we expect that these divergences can be absorbed into the redefinition of the gauge field strength in the propagator , with the form ( [ fren ] ) .
we demonstrate here the calculation of the @xmath114 divergences depicted in fig .
[ loop1 ] . with use of the oscillator representation of the boundary condition for the constant field strength ( [ ( p - fx)b ] ) , @xmath116 we obtain an explicit form of @xmath68 on the boundary as @xmath117
where @xmath118 , and @xmath119 is put to zero ( the boundary value ) later .
then the correlation function is given by @xmath120.\end{aligned}\ ] ] this should be evaluated at the boundary @xmath121 . extracting the short distance behavior ( @xmath122 ) , we have the boundary correlators as @xmath123 there are @xmath124 ways to contract two operators in the vertex @xmath125 $ ] appearing in fig .
[ loop1 ] . noting the correspondence @xmath126 one can see that the contraction ( [ contxx ] ) brings about no divergence after the @xmath127 function regularization . therefore we need only to pick up half ( three ) of the contractions as @xmath128 comparing this result with the propagator in eq .
( [ expand ] ) , the one - loop divergence in the propagator is absorbed into the redefinition of the field strength as @xmath129 this agrees with eq .
( [ fren ] ) , hence the divergences in the boundary state correspond to the quantum loop divergence in the propagator in the string @xmath0 model .
the exponent in the boundary state ( [ bou ] ) consists of the quadratic combination of the oscillators , therefore the boundary state respects the properties of the propagator in the string @xmath0 model .
here we start from the single @xmath0 model action ( [ ssigma ] ) and expand it , hence it is plausible that the divergences of both the propagator and the vacuum graph can be absorbed into the same redefinition of the gauge field .
having defined the boundary state @xmath37 for an arbitrary configuration of the gauge field @xmath7 , our next task in this section is to extract a particular part of the boundary state which describes the interaction with the massless modes of the closed string .
this contributes to the corrections to the d - brane action . as confirmed in ref .
@xcite , the couplings between the gauge field on the d - brane and the closed string field are described in terms of sft as @xmath130 .
thus this interaction ( source term ) is naturally generalized to the one with non - constant field strength : @xmath131 { { \left| b\bigl(f(x)\bigr)\right\rangle}}.\end{aligned}\ ] ] here we have substituted the representation ( [ bcreate ] ) of @xmath41 , therefore now we adopt the approximation considered in the previous section . the first term on the rhs of eq .
( [ braket ] ) gives the same contribution to the part of the d - brane action linear in the closed string massless fields , as the one obtained in the case of the constant field strength @xmath132 in ref .
this is because , after putting the coefficient of the @xmath93 term in eq .
( [ aren ] ) to zero , then we see @xmath133 .
the result is @xcite @xmath134 .
\label{result}\end{aligned}\ ] ] here @xmath135 is the fluctuation of the metric field , @xmath136 is the kalb - ramond two - form field , and @xmath137 is the dilaton field .
the divergent @xmath138 is regularized by the zeta - function regularization .
the resultant quantity ( [ result ] ) is a part of the well known d - brane action linear in the closed string massless fields , , @xmath136 and @xmath139 which exist in the ordinary d - brane action are expected to be reproduced by performing the integrations over the closed string massive modes .
note that thus we obtain a d - brane action of a form _ before _ integrating out the massive modes , _
i.e. _ we can read couplings between the open string gauge field strength @xmath23 and closed string massive modes . ]
@xmath140 the metric in this action is the string metric , model approach , in the case of the constant field strength and the constant tilt angle .
further generalization on general deformation of the d - brane surface will be presented in our next paper @xcite . ] which is related to the fluctuation @xmath141 of the einstein metric through weyl rescaling as @xmath142 the second term in eq .
( [ braket ] ) gives _ finite _ corrections to the d - brane action , that include two derivatives acting on the field strength @xmath94 . in order for the bracket to be expressed in terms of , for example ,
the massless component fields of the closed string , the oscillator contraction should be performed .
explicitly , the second term on the rhs of eq .
( [ bcreate ] ) reads @xmath143 { { \left| b(f)\right\rangle}}}{\nonumber}\\ & & = -{1\over 8}\ ; { \partial}_\rho{\partial}_\delta f_{\mu\nu}\ ; \left [ j^{\kappa\delta}j^{\rho\lambda}j^{\mu\nu } \right ] \alpha^{(-)}_{-1\;\kappa } \alpha^{(+)}_{-1\;\lambda}{{\left| b(f)\right\rangle}}{\nonumber}\\ & & \;\;\;\ ; + \ ; { 1\over 64}\ ; { \partial}_\gamma f_{\alpha\beta}{\partial}_\rho f_{\mu\nu}\ ; \left [ j^{\rho\gamma}j^{\beta\nu}(j^{\mu\alpha}+j^{\alpha\mu } ) \right]{{\left| b(f)\right\rangle}}{\nonumber}\\ & & \;\;\;\ ; + \ ; { 1\over 32}\ ; { \partial}_\gamma f_{\alpha\beta}{\partial}_\rho f_{\mu\nu}\ ; \left [ 2j^{\kappa\mu}j^{\nu\beta}j^{\gamma\rho}j^{\alpha\lambda } -j^{\alpha\mu}j^{\nu\beta}j^{\kappa\rho}j^{\gamma\lambda } \right ] \alpha^{(-)}_{-1\;\kappa}\alpha^{(+)}_{-1\;\lambda}{{\left| b(f)\right\rangle}}{\nonumber}\\[2 mm ] & & \;\;\;\;+\;\cdots\end{aligned}\ ] ] where the matrix @xmath144 is defined as @xmath145 in ( [ finite ] ) we have written explicitly only those terms which contain the closed string massless modes , since these terms are necessary for the evaluation of the d - brane action . the omitted part `` @xmath146 '' in eq .
( [ finite ] ) consists of an infinite number of terms of massive excitations , such as @xmath147 or @xmath148 with @xmath149 .
note that the boundary state for the constant field strength @xmath150 and the massless part of the string field state @xmath151 have even number of @xmath152 oscillators .
( we ignore the auxiliary fields ( denoted as @xmath153 and @xmath154 in ref .
@xcite ) in @xmath28 , since its bracket in eq . ( [ braket ] )
vanishes due to the contraction of the ghost oscillators . )
thus contribution of the terms consisting of odd number of @xmath152 oscillators disappears . with the substitution of eq .
( [ finite ] ) , we find that the second term in eq .
( [ braket ] ) , which is the correction to the d - brane action , equals @xmath155 { \nonumber}\\ & & { \hspace*{-10 mm } } - { t_p\over 256 } \sqrt{-\det ( \eta+f ) } \left [ \left ( h_{\lambda\kappa } + b_{\lambda\kappa } \right)j^{\kappa\lambda } + \left(-4 + { 4\over d-2}j_\lambda^{\;\;\lambda } \right ) d \right ] { \nonumber}\\ & & { \hspace*{60mm}}\times { \partial}_\gamma f_{\alpha\beta}{\partial}_\rho f_{\mu\nu } \left [ j^{\rho\gamma}j^{\beta\nu}(j^{\mu\alpha}+j^{\alpha\mu } ) \right ] .
\label{whresult}\end{aligned}\ ] ] the factor @xmath138 has been regularized in the same way as before .
we have constructed a new boundary state which has the degrees of freedom of an arbitrary configuration of the gauge field @xmath7 on the d - brane .
this state is defined by using the closed sft gauge transformation acting on a conventional boundary state , inspired from the fact that a gauge transformation of a string @xmath0 model generates all modes of the gauge field .
the new boundary state is brst invariant and surely satisfies the nonlinear boundary conditions with the non - constant filed strength .
the new boundary state includes divergences of the order @xmath93 and @xmath138 .
the @xmath138 term can be regularized by the ordinary zeta - function regularization method , whilst the intrinsic divergence of @xmath93 can be absorbed into the redefinition of the gauge field .
this redefinition exactly coincides with the one - loop calculation for the gauge field coupling in a string @xmath0 model . in order for the generalized boundary state to be well - defined ,
the divergent quantity should vanish , hence this constraint forces the gauge field to satisfy a differential equation .
this equation is identical with the conformal invariance condition @xmath156 derived in the string @xmath0 model approach .
the origin of this coincidence is studied in the @xmath0 model side , and we have found that the divergences in the new boundary state can be identified with the one - loop contribution to the worldsheet propagator in the string @xmath0 model . utilizing this boundary state ,
we have computed the derivative correction terms in the d - brane action , by extracting the part corresponding to the massless excitations of the closed string . after putting the divergence in the redefined gauge field to zero
, we have obtained couplings between the derivatives of the field strength ( @xmath114 and @xmath157 ) and the closed string massless fields ( @xmath135 , @xmath136 and @xmath137 ) .
it might be possible to apply the procedure given in ref .
@xcite to obtain a gauge invariant complete sft action .
although this seems to be straightforward , we find it actually complicated . only with a naive extension of the method in ref .
@xcite , there may not be corrected born - infeld action whose variation under the sft gauge transformation cancels the surface term stemming from the source term @xmath158 .
there still remains many subjects to be studied . in this paper
we have argued the non - constant mode of the field strength on the d - brane , but it is more intrinsic to introduce massive modes of the _ open _ strings . though the couplings to the massive modes of the closed strings has been manifested in ref .
@xcite , the boundary state contains infinite number of product of the gauge field , which is a sign of the massive open strings . to construct more general state
, the sft gauge transformation will play an important role .
the closed sft adopted in sec .
[ sec : extending ] has been generalized to a open - closed mixed system @xcite , hence some correspondence between our source term and open - closed coupling in ref .
@xcite may help to understand the boundary state further .
at least , the corrections ( [ whresult ] ) may be reproduced in the low energy sector of the system of ref .
@xcite .
one of the other subjects is concerning the supersymmetry .
unfortunately , in the sft scheme it is hard to discuss the supersymmetry because of some schematic problems @xcite .
a lot of interesting aspects of d - branes are on the basis of the supersymmetry .
we expect that the definition of the new boundary state given in this paper can be generalized to the supersymmetric one .
this sort of generalization is seen in ref .
@xcite , in which the case of constant field strength was treated . with use of the generalized supersymmetric boundary state
, one can calculate corrections to the d - brane actions in the superstring case as well as various interesting quantities @xcite .
* acknowledgments * + i am indebted to h. hata for his collaboration at the early stage of this work .
i would like to thank also a. hashimoto , y. matsuo , and m. akashi for useful comments .
i am grateful to k. furuuchi and n.ishibashi for valuable discussions and careful reading of this manuscript .
this work is supported in part by grant - in - aid for scientific research from ministry of education , science , sports and culture ( # 3160 ) .
i appreciate hospitality of the organizers of summer institute 99 where a part of this work was argued .
* appendix *
in order to view the meaning of the unitary rotational operator @xmath42 $ ] , it is important to check that the definition ( [ bnew ] ) really follows from a gauge rotation ( @xmath19 product ) in sft .
what should be checked is @xmath159 for an infinitesimal @xmath160 ( as a parameter in the @xmath20 ) , instead of the previous @xmath35 . it is needed only to trace the app .
b of ref .
@xcite , for @xmath161 with arbitrary powers in @xmath83 .
we will present only the summary of the calculation . in this case
we should take into consideration the @xmath162 term ( eq.(b.6 ) in @xcite ) @xmath163 in the three string vertex operator @xmath164 .
this @xmath162 factor is quadratic in @xmath3 , thus when treating @xmath165 linear in @xmath83 ( @xmath166 ) as in @xcite its contribution disappears .
but for higher powers in @xmath83 , the logarithmic divergence @xmath167 comes out .
this divergences are absorbed into the normal ordering divergences of the @xmath68 s in @xmath168 , but its finite part remains ambiguous .
for example , the gauge transformation parameter @xmath169 results in the identity @xmath170 { { \left| b(f)\right\rangle } } , \label{c}\end{aligned}\ ] ] where @xmath171 is a _
finite _ unknown constant ( contractions of @xmath172 comes from the flat metric @xmath173 in eq .
( [ fp ] ) ) .
in this paper we assume that @xmath174 , since if we define @xmath41 with non - zero @xmath171 using the above eq .
( [ c ] ) , the generalized nonlinear boundary condition ( [ eigen ] ) does not follow .
after all , together with the above results and the definitions ( [ transnew ] ) ( [ bnew ] ) , the gauge transformation of the dynamical variable @xmath7 of the new boundary state @xmath41 generated by the sft @xmath19 product is ( [ gaugetr ] ) , for an arbitrary number of powers in @xmath175 .
in this appendix we present the calculation to obtain eq.([bcreate ] ) , changing all the oscillators into creation operators using the boundary condition ( [ bc ] ) .
we show here the calculation of only the last term in eq .
( [ v ] ) , since this term is the most bothersome and other terms can be evaluated in the same way . considering the corresponding mode of the gauge field @xmath176 the last term in eq .
( [ v ] ) is written as @xmath177 { \hspace*{20mm}}{\nonumber}\\ \times \left [ \sum_{k , l}\frac{1}{kl } \left(\alpha_{-k - l}^{(+)\alpha}-\alpha_{k+l}^{(-)\alpha } \right ) \left(\alpha_{k}^{(+)\beta}-\alpha_{-k}^{(-)\beta } \right ) \left(\alpha_{l}^{(+)\gamma}-\alpha_{-l}^{(-)\gamma } \right ) \right].\end{aligned}\ ] ] this is evaluated on the boundary state @xmath87 . after changing all the oscillators into creation operators , then we are led to the following form symbolically : @xmath178 { { \left| b\bigl(f(x)\bigr)\right\rangle}}. \label{a6}\end{aligned}\ ] ] on the rhs , the first two terms have finite coefficients and furthermore are not relevant for the massless modes in the closed string excitations , hence it is not necessary to evaluate them explicitly .
the third term quadratic in @xmath152 and the last constant term @xmath171 contain divergences and also finite contributions .
the @xmath179 term is given as @xmath180{\nonumber}\\ & + \displaystyle\sum_{m>0 , n>0 } \alpha^{(-)\sigma}_{-n - m}\alpha^{(+)\delta}_{-n - m } \biggm [ & \frac{-2}{m(n\!+\!m ) } ( -j^{\sigma\mu}j^{\nu\alpha}j^{\rho\beta}j^{\gamma\delta } -j^{\mu\delta}j^{\alpha\nu}j^{\beta\rho}j^{\sigma\gamma } ) { \nonumber}\\ & & + \frac{-2}{n(n\!+\!m ) } ( -j^{\sigma\mu}j^{\nu\beta}j^{\rho\alpha}j^{\gamma\delta } -j^{\mu\delta}j^{\beta\nu}j^{\alpha\rho}j^{\sigma\gamma } ) { \nonumber}\\ & & \;\;+\frac{2}{mn } ( -j^{\sigma\mu}j^{\nu\beta}j^{\rho\gamma}j^{\alpha\delta } -j^{\mu\delta}j^{\beta\nu}j^{\gamma\rho}j^{\sigma\alpha } ) \biggm ] { \nonumber}\\ & + \;\;\mbox{other summations .
} & \label{summ}\end{aligned}\ ] ] here `` other summations '' denotes the summation in the other four regions in the @xmath181-plane divided by three lines : @xmath182 .
the other summation is performed in the same manner with slight changes of the indices . from this expression
we can extract the divergences @xmath93 .
for example , from the first summation we have @xmath183,\end{aligned}\ ] ] where we have defined @xmath184 .
this is because , in this case we should sum over @xmath185 as the index of the oscillators , and for a fixed @xmath185 , divergent quantities proportional to @xmath93 stem from the first and the second terms in the first summation in eq .
( [ summ ] ) ( the factor of the third term @xmath186 does not give the divergence ) .
in the same way , one finds that the second summation in eq .
( [ summ ] ) does not bring about the divergence .
collecting all the divergent terms , we can verify the following equality : @xmath187 with @xmath188-(\kappa\leftrightarrow\lambda ) .
\label{deltaf}\end{aligned}\ ] ] in the same way , the constant term @xmath171 in eq .
( [ a6 ] ) is evaluated , and the result is @xmath189 with the above same @xmath190 ( [ deltaf ] ) .
these give the relation ( [ bcreate ] ) , using the identification @xmath191 s. p. de alwis and k. sato , _ phys .
* d53 * ( 1996 ) 7187 hep - th/9601167 ; + a. a. tseytlin , _ nucl .
_ * b469 * ( 1996 ) 51 , hep - th/9602064 ; + y. lozano , _ phys .
_ * b399 * ( 1997 ) 233 , hep - th/9701186 ; + m. aganagic , j. park , c. popescu and j. h. schwarz , _ nucl .
phys . _ * b496 * ( 1997 ) 215 , hep - th/9702133 ; + y. igarashi , k. itoh , k. kamimura and r. kuriki , _ jhep _ * 9803 * ( 1998 ) 002 , + hep - th/9801118 ; + i. oda , _ phys .
lett . _ * b430 * ( 1998 ) 242 , hep - th/9802152 ; + y. igarashi , k. itohand k. kamimura , _ nucl .
_ * b536 * ( 1998 ) 469 , hep - th/9806161 . | in this paper , we generalize a boundary state to the one incorporating non - constant gauge field strength as an external background coupled to the boundary of a string worldsheet in bosonic string theory .
this newly defined boundary state satisfies generalized nonlinear boundary conditions with non - constant gauge field strength , and is brst invariant .
the divergence immanent in this boundary state coincide with the one calculated in a string @xmath0 model .
we extract the relevant massless part of this generalized boundary state , and give a part of the d - brane action with the non - constant gauge field strength , that is , derivative corrections to the d - brane action .
22.5cm16.8cm-.4cm-.9 cm = 6pt plus 2pt minus 1pt plus 0.2pt minus 0.1pt bgnmark < endmark > bgncnt endcnt h@ight @exist tempdimc tempdimd @ight @dth # 1<#2#3#4@mae#1mark#2 front#3usiro#4 # 1>#2#3#4@mae#1mark#2 front#3usiro#4 @lc 0= tempdima0 tempdimb0 tempdima by2 tempdima by tempdimc tempdima by tempdimd dblfornoop#1@#2#3#4 dblfor#1;#2:=#3#4fortmp#3fortmpempty dblforloop#3nil , nil , nil@#1#2#4 dblforloop#1,#2,#3@#4#5#6#4#1 # 5#2 # 4nnil nextwhile = dblfornoop # 6nextwhile = dblforloopnextwhile#3@#4#5#6 @endpt#1#2dblfor;:=#2@h@ight#1#2 tempcnta@ tfor:=#2 # 1#2#3#4#5 tempdima=#3 tempdima-#1 tempdimc=#5@ight tempdimb = tempdimc tempdimb-@dth tempdimd=#2 tempdimd-1.587ex # 1 height tempdimc width@dth depth - tempdimd -@dth height tempdimc widthtempdima depth - tempdimb-@dth tempdimd=#4 tempdimd-1.587ex height tempdimc width@dth depth - tempdimd # 1mae#1nil # 1ato#1nil mae#1;#2nil#1 ato#1;#2nil#2 # 1#2 @ight = ex @dth = em wickdata @list bgncnt@ endcnt@ list @sition @sition @ng#2 tfor@mber:=#2 listlistusiro bgncnt
= endcnt latexerrthe numbers of ` < ' and ` > ' do not match you have written different numbers of < and > @ tfor:=#1 bgncnt= latexerrthe number of numbers in the first argument is different with that of contractions
< ... > give the same numbers of heights as the contractions < ... > addtoresetequationsection |
non - abelian gauge theories like qcd are known to possess a rich vacuum structure .
notably , there are topologically non - trivial vacuum fluctuations of the gauge fields , whose simplest examples are instantons @xcite .
instantons have been argued to play an important role in various long - distance aspects of qcd , such as giving a possible solution to the axial @xmath0 problem @xcite or being at work in @xmath1 chiral symmetry breaking @xcite .
recently , also a number of qcd lattice studies elucidated the topological structure and the instanton content of the vacuum @xcite .
another prominent feature of the qcd vacuum is its apparent dual superconductivity .
the latter has been recognized since long as an explanation of confinement @xcite and is also intensively studied on the lattice @xcite .
these investigations seem to indicate that the fundamental dual excitations are monopoles or vortices , while instantons might be composite objects of those .
the qcd vacuum structure may be studied on the lattice , starting from first principles .
alternatively , low - energy phenomenology can be exploited to learn about the vacuum , at the expense of introducing model assumptions as for example in the instanton liquid model @xcite or in the model of the stochastic vacuum @xcite , which incorporates the dual superconductor picture of the vacuum . in this short review
, i shall emphasize that manifestations of the vacuum structure of non - abelian gauge theories can also be searched for in high - energy scattering .
i shall concentrate in sect .
[ ihard ] on instanton - induced hard scattering processes in the electroweak gauge theory ( quantum flavor dynamics ( qfd ) ) @xcite and in deep - inelastic scattering ( dis ) in qcd @xcite , which are calculable from first principles within instanton - perturbation theory . in sect .
[ soft ] , i shall discuss briefly soft scattering processes in qcd .
instantons ( @xmath2 ) are minima of the euclidean action , localized in space and euclidean time , with unit topological charge @xmath3 . in minkowski space - time , they describe tunneling transitions between classically degenerate , topologically inequivalent vacua , differing in their winding number by one unit , @xmath4 @xcite . the corresponding energy barrier , under which the instantons tunnel , is inversely proportional to the gauge coupling @xmath5 and the effective @xmath2-size @xmath6 , @xmath7 , and of order @xmath8 tev in qfd @xcite and @xmath9 in hard scattering in qcd , where @xmath10 gev is a large momentum transfer e.g. in dis @xcite . due to axial anomalies @xcite ,
@xmath2-induced hard scattering processes are always associated with anomalous fermion - number violation @xcite , in particular baryon plus lepton number violation , @xmath11 , in the case of qfd , chirality violation , @xmath12 , in the case of qcd .
generically , @xmath2-induced total cross - sections for hard parton scattering processes are given in terms of an integral over the instanton - antiinstanton ( @xmath13 ) collective coordinates ( sizes @xmath14 , @xmath13 distance @xmath15 , color orientation @xmath16 ) @xcite ( see also @xcite ) @xmath17 here , the basic blocks arising in @xmath2-perturbation theory are the @xmath2-size distribution @xmath18 and the function @xmath19 , which takes into account the exponentiation of gauge boson production @xcite and can be identified with the @xmath13-interaction defined via the valley method @xcite . the @xmath2-size distribution @xmath18 is known in @xmath2-perturbation theory , @xmath20 , up to two - loop renormalization group invariance @xcite
. at one - loop
, it reads @xmath21 with @xmath22 the first coefficient in the @xmath23 function , @xmath24 the @xmath2-action , @xmath25 the renormalization scale , and @xmath26 a scheme - dependent constant .
the quite differenct @xmath27 dependence of the size distribution ( [ size - dist ] ) for qcd and qfd has important consequences for the predictivity : whereas large - size instantons , @xmath28 , are exponentially suppressed in qfd ( cf .
( [ action ] ) ) and thus the relevant contributions to the size integrals in ( [ gencross ] ) arise consistently from the perturbative region ( @xmath29 ) even if both initial partons are on - shell , @xmath30 , in qcd the power - law behavior of the size distribution , @xmath31 , generically causes the dominant contributions to ( [ gencross ] ) to originate from large @xmath32 and thus often spoils the applicability of @xmath2-perturbation theory . in dis ( cf.fig.[kin-var ] ) , however , one parton , @xmath33 say , carries a space - like virtuality @xmath34 , such that the contribution of large instantons to the integrals is suppressed by an exponential factor in ( [ gencross ] ) , @xmath35 , and @xmath2-perturbation theory becomes exploitable , i.e. predictable , for sufficiently large @xmath36 @xcite . in this connection
it is quite welcome that lattice data on the instanton content of the quenched ( @xmath37 ) qcd vacuum @xcite can be exploited to infer the region of validity of @xmath2-perturbation theory for @xmath18 @xcite : as illustrated in fig .
[ lattice ] ( left ) , there is very good agreement for @xmath38 fm .
[ h ] + as far as the @xmath13-interaction @xmath19 is concerned , it is seen from simple fourier correspondences in ( [ gencross ] ) , e.g. @xmath39 , that at high center - of - mass ( cm ) energies @xmath40 small @xmath13-distances are probed , i.e. @xmath41 ( dis ) or @xmath42 ( qfd ) .
thus , in order to make a reliable prediction of @xmath2-induced hard scattering at high energies , we need to know the interaction for small distances . again , for qcd one may exploit lattice data on the @xmath13-distance distribution @xcite to infer the region of validity of the description of the @xmath13-interaction by its exact expression given in the valley method @xcite : one finds good agreement for @xmath43 @xcite ( cf .
[ lattice ] ) . in this case
, however , there are remaining theoretical ambiguities : the integrations over @xmath27 , @xmath44 in the @xmath13-distance distribution imply significant contributions also from larger instantons with @xmath45 fm , outside the strict region of perturbation theory .
it is not excluded that the valley interaction reliably describes the interactions of small - size instantons , @xmath46 fm , at smaller @xmath47 , say @xmath48 .
[ h ] the energy dependencies of the @xmath2-induced parton - parton cross sections illustrated in fig .
[ isorho ] for dis ( @xmath49 ) @xcite and in fig .
[ vlhc ] for qfd @xcite are easily understood on the basis of above mentioned fourier correspondences and associated saddle - point relations : at increasing energies , smaller and smaller @xmath13-distances are probed , and the cross - sections are rapidly growing due to the attractive nature of the @xmath13-interaction in the perturbative semi - classical regime ( cf .
[ lattice ] ( right ) ) .
this goes along with the production of @xmath50 gauge bosons , in addition to the minimal number of fermions required by the axial anomaly .
furthermore , in the case of dis , at increasing virtuality @xmath51 smaller and smaller instantons are probed and the cross - section diminishes in accord with the size distribution ( fig.[lattice ] ( left ) ) . with my colleague fridger schrempp
we have conducted a long - term research program at desy to work out the theory and phenomenology of hard qcd @xmath2-induced processes in dis at hera @xcite .
meanwhile , the first dedicated search by the h1 collaboration has been published @xcite .
several observables characterising the hadronic final state of qcd @xmath2-induced events were exploited to identify a potentially @xmath2-enriched domain .
the results obtained are intriguing but non - conclusive .
while an excess of events with @xmath2-like topology over the expectation of the standard dis background is observed , which , moreover , is compatible with the @xmath2-signal , it can not be claimed to be significant given the uncertainty of the monte carlo simulations of the standard dis background .
the data do not exclude the cross - section predicted by @xmath2-perturbation theory for quite small @xmath2-sizes @xmath52 fm and @xmath53 .
[ h ] in the early
90s the possibility of observable qfd @xmath2-effects at cm energies @xmath54 tev was quite intensively investigated @xcite . but despite considerable theoretical efforts the actual size of the cross - sections in the relevant energy regime was never established . in view of the similarity between qfd and hard qcd @xmath2-induced processes in dis and of the recent information about the latter both from the lattice and from experiment , it seems worthwhile to reconsider the subject @xcite .
figure [ vlhc ] represents a state - of - the - art evaluation of the qfd @xmath2-induced quark - quark cross - section , which neglects higgs production @xcite .
it demonstrates that the cross - section is unobservably small in the conservative fiducial region corresponding to @xmath55 .
however , it becomes of observable size if the @xmath13-attraction remains valid also for slightly smaller values of @xmath56 at @xmath57 tev .
this opens up exciting opportunities at future colliders @xcite such as vlhc @xcite or at cosmic ray facilities and neutrino telescopes @xcite .
one of the main challenges of qcd is to understand the high - energy , but small momentum transfer cross - sections of hadrons from first principles . in this domain ,
perturbation theory is not applicable and direct lattice simulations are not possible until now .
there have been recently several attempts to face this challenge @xcite .
most of them are based on the eikonal representation of the amplitudes in terms of correlators of wilson lines or loops @xcite , which thereafter are evaluated using different model assumptions .
the most advanced attempt is certainly the one of the heidelberg group @xcite based on the model of the stochastic vacuum @xcite , which has been already successfully applied to a variety of soft high - energy processes ( e.g. @xcite ) . less developed but
rapidly evolving is the understanding of a possible connection of semi - hard instantons to diffraction @xcite .
there exists the intriguing possibility that larger - size instantons build up diffractive scattering , with the marked @xmath2-size scale @xmath58 fm ( fig .
[ lattice ] ( left ) ) being reflected in the conspicuous `` geometrization '' of soft qcd @xcite .
a. belavin _ et al .
_ , phys .
b * 59 * ( 1975 ) 85 .
g. t hooft , phys .
d * 14 * ( 1976 ) 3432 .
e. v. shuryak , nucl .
b * 203 * ( 1982 ) 93 ; d. diakonov and v. y. petrov , nucl .
b * 272 * ( 1986 ) 457 ; t. schfer and e. v. shuryak , rev .
* 70 * ( 1998 ) 323 .
m. teper , nucl .
* 83 * ( 2000 ) 146 .
g. t hooft , in _ high energy physics _ , eps international conference , palermo 1975 ; ed .
a. zichichi ; s. mandelstam , phys .
* 23 * ( 1976 ) 245 .
a. di giacomo , in _ confinement , topology , and other nonperturbative aspects of qcd _ , nato advanced research workshop , stara lesna , slovakia , 2002 , arxiv : hep - lat/0204032 . h. g. dosch and y. a. simonov , phys .
b * 205 * ( 1988 ) 339 .
a. ringwald , nucl .
b * 330 * ( 1990 ) 1 ; o. espinosa , nucl .
b * 343 * ( 1990 ) 310 .
a. ringwald , f. schrempp and c. wetterich , nucl .
b * 365 * ( 1991 ) 3 ; m. gibbs _ et al .
_ , z. phys .
c * 66 * ( 1995 ) 285 .
v. i. zakharov , nucl .
b * 371 * ( 1992 ) 637 .
m. porrati , nucl .
b * 347 * ( 1990 ) 371 ; v. v. khoze and a. ringwald , nucl .
b * 355 * ( 1991 ) 351 .
d. a. morris and a. ringwald , astropart .
* 2 * ( 1994 ) 43 .
v. a. rubakov and m. e. shaposhnikov , usp . fiz .
nauk * 166 * ( 1996 ) 493 .
a. ringwald and f. schrempp , in _
quarks 94 _ , eighth international seminar , vladimir , russia , 1994 , arxiv : hep - ph/9411217 . s. moch , a. ringwald and f. schrempp , nucl . phys .
b * 507 * ( 1997 ) 134 .
a. ringwald and f. schrempp , phys .
b * 438 * ( 1998 ) 217 .
a. ringwald and f. schrempp , phys .
b * 459 * ( 1999 ) 249 .
a. ringwald and f. schrempp , comput .
phys . commun .
* 132 * ( 2000 ) 267 .
a. ringwald and f. schrempp , phys .
b * 503 * ( 2001 ) 331 .
r. jackiw and c. rebbi , phys .
* 37 * ( 1976 ) 172 ; c. g. callan , r. f. dashen and d. j. gross , phys .
b * 63 * ( 1976 ) 334 .
f. r. klinkhamer and n. s. manton , phys .
d * 30 * ( 1984 ) 2212 . s. l. adler , phys .
* 177 * ( 1969 ) 2426 ; j. s. bell and r. jackiw , nuovo cim .
a * 60 * ( 1969 ) 47 .
i. balitsky and v. m. braun , phys .
d * 47 * ( 1993 ) 1879 ; phys .
b * 314 * ( 1993 ) 237 .
a. v. yung , nucl .
b * 297 * ( 1988 ) 47 .
j. verbaarschot , nucl .
b * 362 * ( 1991 ) 33 . c. w. bernard , phys .
d * 19 * ( 1979 ) 3013 ; t. r. morris , d. a. ross and c. t. sachrajda , nucl .
b * 255 * ( 1985 ) 115 .
d. a. smith and m. j. teper [ ukqcd collaboration ] , phys .
d * 58 * ( 1998 ) 014505 .
a. ringwald , desy 02 - 157 .
a. hebecker , in _
hep 2001 _ , international europhysics conference on high - energy physics , budapest , hungary , 2001 , arxiv : hep - ph/0111092 .
o. nachtmann , annals phys .
* 209 * ( 1991 ) 436 .
h. g. dosch , e. ferreira and a. krmer , phys .
d * 50 * ( 1994 ) 1992 ; o. nachtmann , in _ workshop on topics in field theory _ , kloster banz , germany , 1993 , arxiv : hep - ph/9609365 . h. g. dosch , o. nachtmann , t. paulus and s. weinstock , eur .
j. c * 21 * ( 2001 ) 339 .
a. i. shoshi , f. d. steffen and h. j. pirner , nucl .
phys . a * 709 * ( 2002 ) 131 .
d. e. kharzeev , y. v. kovchegov and e. levin , nucl .
a * 690 * ( 2001 ) 621 ; m. a. nowak , e. v. shuryak and i. zahed , phys .
d * 64 * ( 2001 ) 034008 ; f. schrempp , j. phys .
g * 28 * ( 2002 ) 915 .
f. schrempp and a. utermann , phys .
b * 543 * ( 2002 ) 197 . | this short review deals with the manifestations of the vacuum structure of non - abelian gauge theories in high - energy scattering .
specifically , it concentrates on instanton - induced hard scattering processes , both in the electroweak gauge theory and in qcd .
soft scattering processes in qcd and their connection to models of semi - hard vacuum fluctuations are also briefly discussed . |
flavor models based on discrete symmetries are often used to address issues like large / small mixing and mild / strong mass hierarchies in the lepton / quark sector ( for reviews , see , e.g. @xcite ) .
the permutation group @xmath0 is an attractive such candidate which was introduced in @xcite and explored further in @xcite . in this paper
we study the exciting prospect that such flavor models can predict _ enlarged higgs sectors with non - standard couplings to fermions and gauge bosons _ , although such a nonsupersymmetric extension does not provide any additional stability to the potential .
the main motivation here arises from flavor issues .
its supersymmetrization would indeed bring in quantum stability , but in this work we stick to a minimal nonsupersymmetric @xmath0 flavor scenario .
the motivation for choosing @xmath0 is that it is the smallest non - abelian discrete symmetry group that contains a irreducible representation which can connect two maximally mixed generations .
it is the symmetry group of an equilateral triangle and has three irreducible representations : @xmath4 and @xmath5 , with multiplication rules : @xmath6 and @xmath7 . besides facilitating maximal mixing through its doublet representation ,
@xmath0 provides two inequivalent singlet representations which play a crucial role in reproducing fermion masses and mixing . to accomplish the latter , three scalar @xmath1 doublets are introduced , which couple to the fermions as dictated by @xmath0 symmetry .
it so happens that large mixing among up- and down - type quarks cancel each other in the cabibbo matrix .
neutrino masses are separately generated by a type - ii see - saw mechanism using scalar @xmath1 triplets @xcite , so that the mismatch between the large mixing of the charged leptons and the diagonal neutrino masses translates directly into the pontecorvo - maki - nakagawa - sakata matrix . in this paper
we do not deal with those triplets , but explore the following avenues : ( @xmath8 ) minimization of the scalar potential with three scalar @xmath1 doublets , two of which form an @xmath0 doublet and the third an @xmath0 singlet , ( @xmath9 ) the gauge and yukawa interactions of the neutral scalars , and ( @xmath10 ) different nonstandard production and decay modes of the neutral cp - even scalars leading to the possibility of their detection at the lhc . for definiteness , we study the @xmath0 model pursued in @xcite to explain the leptonic flavor structure .
we concentrate on the complementary aspects by exploring the scalar sector .
the assignments of the fermion and scalar fields are as follows : @xmath11 where the notations are standard and self - explanatory .
the vacuum expectation values ( vevs ) of the three scalar doublets @xmath12 induce spontaneous electroweak symmetry breaking ( ssb ) .
the most general @xmath0 invariant scalar potential involving three scalar doublet fields is given by @xcite @xmath13 \ , .\end{gathered}\ ] ] after ssb , nine degrees of freedom are left : three neutral scalars , two neutral pseudoscalars and two charged scalars with two degrees of freedom each .
we denote the vevs of @xmath14 by @xmath15 and assume the @xmath16 s to be real . for the purpose of generating maximal mixing in the lepton sector , we choose the vacuum alignment @xmath17 .
once we solve the tadpole equations , @xmath18 turns out to be an extremal condition if the following relations are satisfied : @xmath19 to ensure that the chosen vacuum alignment actually corresponds to a minimum of the potential , we adjust parameters to make sure that the determinant of the hessian matrix is positive .
just to recall , the hessian is defined as the square matrix of second order partial derivatives of a function describing its local curvatures . in this case , the function is the scalar potential and the hessian is just the mass matrix of the scalars .
the positivity of the eigenvalues see later in eq .
( [ eqn : scalarmasses ] ) guarantees that the potential is minimized .
we note that after ssb , the potential turns out to be a polynomial of order four , and its global stability in the asymptotic limit ( i.e. @xmath20 ) is ensured by the following set of conditions : @xmath21 we now set out to find the spectrum of the three cp - even neutral scalars .
we insert the expansion @xmath22 in eq .
( [ eqn : scalarpotential ] ) to obtain the mass matrix . after its diagonalization
the weak basis scalars @xmath23 are expressed in terms of the physical scalars @xmath24 as @xmath25 where @xmath26 and @xmath27 are analytically tractable but complicated functions of @xmath16s , @xmath28 and @xmath29 , which we do not display .
the condition @xmath30 immediately leads to @xmath31 and @xmath32 .
the masses of the three cp - even neutral scalars are @xmath33\ , ; \label{eqn : mabc}\end{aligned}\ ] ] where @xmath34^\frac{1}{2}\ , .\end{aligned}\ ] ] a few things are worth noting at this stage : 1 .
since @xmath12 are all weak @xmath1 doublets , their vevs are related as : @xmath35 , where @xmath36 .
one of the physical scalars is given by @xmath37 , i.e. there is no dependence on @xmath38 or on the vevs .
this happens because @xmath0 symmetry requires the scalar mass matrix to be of the form @xmath39 which always yields @xmath40 as one eigenvector , regardless of the values of @xmath41 and @xmath42 .
we strictly follow eq .
( [ min_conditions ] ) to ensure that the potential is bounded from below .
we randomly vary the _ magnitude _ of the @xmath16 s in the range @xmath43 $ ] , although slightly larger ( but @xmath44 ) values of @xmath45 would have still kept the couplings perturbative .
accept _ a given set of @xmath46 only if it satisfies the minimization conditions .
the difference @xmath47 is positive , and when @xmath48 , i.e. @xmath49 , the splitting grows enormously .
since the maximum value of @xmath50 is controlled by @xmath51 , @xmath52 becomes tachyonic when @xmath48 .
it has been suggested in @xcite that with _ order one _ yukawa couplings , the ratio @xmath53 reproduces the correct cabibbo angle in the quark sector .
we require @xmath54 to ensure that @xmath52 stays above the accepted limit . since @xmath55 and @xmath56 have similar gauge and yukawa properties , quite different from those of @xmath57 ( see discussions later ) , we show the mass splitting @xmath58 against @xmath59 in fig . [ fig : scalar_massrelations](a ) , and the relation between @xmath59 and @xmath60 in fig .
[ fig : scalar_massrelations](b ) .
the kinetic terms @xmath61 ( @xmath62 ) yield the couplings of the symmetry basis @xmath63 to @xmath64 and @xmath65 . clearly , these couplings are modified by a factor of @xmath66 compared to their sm expressions . in terms of the mass basis scalars ,
we observe the following : ( @xmath8 ) the coupling of @xmath55 to @xmath67 ( or , @xmath68 ) is the corresponding sm coupling multiplied by @xmath69 and the corresponding factor for @xmath56 is @xmath70 .
( @xmath9 ) the scalar @xmath57 does not have @xmath71 or @xmath72 couplings , unlike the other two scalars .
this can be understood as follows .
the gauge couplings of @xmath63 arise from the linear expansion @xmath73 in the kinetic term . since @xmath74
, the combination @xmath75 will couple to gauge bosons as proportional to @xmath28 . the orthogonal combination @xmath76 that represents the physical scalar @xmath57 see eq .
( [ eqn : scalarmixing ] ) and point @xmath9 following eq .
( [ eqn : delta - m3 ] ) will not have the usual scalar - gauge - gauge vertex .
the four - point @xmath77 and @xmath78 couplings will , however , exist .
the @xmath0 invariant yukawa lagrangian , where the neutral scalars and the charged leptons / quarks are in their weak basis , is given by ( @xmath79 s are the leptonic and @xmath80s are quark yukawa couplings ) @xmath81 the couplings of @xmath82 to the quarks and leptons depend on the parameters @xmath83 and @xmath79 ( or , @xmath80 ) , while the couplings of @xmath57 to fermions depend only on @xmath79 ( or , @xmath80 ) .
the _ physical _ scalar couplings to the _ mass basis _
fermions are given by the following yukawa matrices , displayed for the charged leptons as an example ( the structures for the quark sector are similar _ modulo _ cabibbo mixing ) : @xmath84 the position of the zeros in the matrices deserves some attention .
it turns out that @xmath24 have off - diagonal fermion couplings at tree level due to the absence of any natural flavor conservation @xcite .
the numerical entries of the yukawa matrices are intimately tied to the successful reproduction of the quarks and leptons masses and mixings .
we make three observations at this stage .
@xmath85 we admit that unless some @xmath0 breaking parameters are introduced a successful reproduction of @xmath86 and @xmath87 is problematic @xcite , and also domain walls will be formed .
we do not intend to cover all flavor issues .
we simply concentrate on the scalar sector whose lagrangian is @xmath0 invariant to start with .
@xmath57 couples _ only _ off - diagonally and one of the two fermions has to be from the third generation .
@xmath89 @xmath82 couple diagonally as in the sm , _ but also _ possess small , numerically insignificant , off - diagonal couplings involving the first two generations .
the last two points require further clarification . in a theory with more than one @xmath1 higgs doublet ,
tree level flavor changing neutral currents ( fcnc ) generally exist in the scalar sector .
for example , they exist in the ordinary ` two higgs doublet model ( 2hdm ) ' , but in the supersymmetric standard model they are avoided by the arrangement that one doublet couples only to the up - type fermions and the other to only the down - types . in nonsupersymmetric scenarios , in the absence of any natural flavor conservation , symmetry arguments have been advanced in the context of multi - higgs models to show that the off - diagonal yukawa couplings of the neutral scalars are suppressed by their relation to the off - diagonal entries of the cabibbo - kobayashi - maskawa ( ckm ) matrix @xcite . in the present case , @xmath0 symmetry , under which both scalars and fermions transform nontrivially ,
is instrumental in suppressing the off - diagonal couplings . to provide intuitive understanding , we take , as an example , only the two - flavor @xmath90@xmath91 sector together with two neutral scalars @xmath92 and @xmath93 .
it is not difficult to see that the combination @xmath94 , which corresponds to @xmath57 , couples only off - diagonally , as mentioned earlier .
but the other combination @xmath95 , which corresponds to @xmath82 following eq .
( [ eqn : scalarmixing ] ) , couples only diagonally to physical @xmath90 or @xmath91 .
when we consider the quark sector , @xmath90 and @xmath91 would be replaced by second and third generation quarks which will have ckm mixing .
this will yield off - diagonal entries for @xmath82 couplings to quarks suppressed by the off - diagonal ckm elements .
the same happens for off - diagonal couplings involving the first two generations as well .
the tiny size of tree level fcnc rates in an @xmath0 flavor model has been noticed also earlier , where predictions for @xmath96 , @xmath97 and @xmath98 have been given @xcite . in some setups where the fermion transformations under @xmath0 are not appropriately adjusted
, the off - diagonal yukawa couplings may become order one which induce sizable neutral scalar mediated rare processes , like @xmath99 or @xmath100 , at tree level .
this requires those neutral scalars to lie beyond several tev @xcite .
but in our case , once we adjust the @xmath101 s to reproduce the fermion masses and mixing , the off - diagonal yukawa couplings are determined too .
the largest of them corresponds to @xmath102 , which is about 0.8 .
the second largest off - diagonal coupling is that for @xmath103 , and is about 0.02 .
the next in line is @xmath104 , whose coefficient is about 0.008 .
the others are orders of magnitude smaller , and are of no numerical significance .
although fcnc processes like @xmath105@xmath106 and @xmath107@xmath108 mixings proceed at tree level , the contributions are adequately suppressed even for light scalar mediators .
the perturbativity condition @xmath109 and the requirement @xmath110gev ( for which we set @xmath111 ) yields @xmath59 in the neighbourhood of 120gev and @xmath112 within 400gev see the scatter plots in fig .
[ fig : scalar_massrelations ] . both @xmath55 and @xmath56
would decay into the _ usual
_ @xmath68 , @xmath113 , @xmath114 , @xmath115 , @xmath116 modes , but the dominant decay mode of @xmath55 ( or @xmath56 ) for the case of @xmath117 ( or @xmath118 ) would be into @xmath119 .
recall that the existing limits on the higgs mass depend crucially on the gauge coupling of the higgs .
since @xmath71 or @xmath72 couplings are nonexistent , the mass of @xmath57 is unconstrained , i.e. @xmath60 can be lower than 114gev or larger than 200gev .
we numerically calculate the strength of the @xmath120 coupling from the set of _ acceptable _ parameters characterizing the potential , and introduce a parameter @xmath121 which is the ratio of the @xmath120 coupling and the @xmath122 coupling .
the magnitude of @xmath121 depends on the choice of @xmath16 and @xmath29 .
assuming @xmath123gev , we obtain @xmath121 in the range of @xmath124 . just to compare with a 2hdm @xcite for illustration , the corresponding @xmath121 value , when the heavier higgs weighing around 400gev decays into two lighter higgs weighing 114gev each , is about 10 . in fig .
[ fig : decays](a ) we have plotted the branching ratio of @xmath125 as a function of @xmath59 for two representative values @xmath123,@xmath126gev , and for @xmath127 5 and 30 , which correspond to the smallest and largest @xmath121 obtained from the set of _ accepted _ scalar parameters .
we observe that till the @xmath113 or @xmath68 decay modes open up , the branching ratio @xmath128 is almost 100% . to calculate the decay widths into the usual modes ( other than @xmath129 ) ,
we have used ` hdecay ` @xcite by appropriately modifying the gauge and yukawa couplings . as fig .
[ fig : decays](b ) suggests , as long as @xmath130 , @xmath57 will dominantly decay into jets , and one of them can be identified as the @xmath131-jet
. the branching ratio of @xmath132 is , nevertheless , not negligible ( about 0.1 ) . as shown in fig .
[ fig : decays](c ) , for @xmath133 , the branching ratio of @xmath134 is quite sizable , which falls with increasing @xmath60 .
it may be possible to reconstruct @xmath57 from @xmath135 .
in fact , a light @xmath57 would be copiously produced from the top decay at the lhc .
on the other hand , if @xmath136 , as can be seen again from fig .
[ fig : decays](b ) , @xmath57 decays to @xmath137 with an almost 100% branching ratio .
if @xmath121 is large , then there is an interesting twist to the failed higgs search at lep-2 . in this case
, @xmath125 would overwhelm @xmath138 , and hence the conventional search for the sm - like scalar ( @xmath55 , as the lighter between @xmath55 and @xmath56 ) would fail .
this is similar to what happens in the next - to - minimal supersymmetric models , when the lightest scalar would dominantly decay into two pseudoscalars , and each pseudoscalar would then decay into @xmath139 or @xmath140 final states . in view of these possible
@xmath141 or @xmath142 higgs signals , lep data have been reanalyzed putting constraints on the higgs production cross section times the decay branching ratios @xcite .
the possibility of the higgs cascade decays into @xmath143 ( @xmath144 quark / gluon ) , @xmath145 photons and 4 photons has been studied too @xcite . from a study of @xmath141 final states ,
a limit @xmath146 gev ( for a sm - like higgs ) has been obtained @xcite . from all other
cascade decays the limit on @xmath147 will be considerably weaker .
our @xmath57 has the special feature that it has only off - diagonal yukawa couplings involving one third - family fermion .
if @xmath55 is lighter than the top quark , it would decay as @xmath148 , and into @xmath149 , the latter constituting a spectacular signal with two different lepton flavors @xmath90 and @xmath91 . the standard @xmath139 and cascade @xmath141 decay searches are not sensitive to our final states , and so a value of @xmath59 much lighter than 110 gev is not ruled out .
the discrete flavor symmetry @xmath0 , besides successfully reproducing fermion masses and mixing , provides an extended higgs sector having unconventional decay properties .
we assume all the couplings to be real , and do not deal with the possibility of cp violation in this paper .
the potential has been minimized requiring maximal mixing for the atmospheric neutrinos . in our setup , there are two scalars which are sm higgs like , _ except _ that each of them can have a dominant decay into the third ( @xmath150 ) .
the latter , i.e. @xmath57 , has no @xmath151-type gauge interactions , and has _ only _ flavor off - diagonal yukawa couplings with one fermion from the third generation .
it is not unlikely that by evading the conventional search strategies , both @xmath55 and @xmath57 are already buried in the existing lep and tevatron data . in this analysis
we have not dealt with the two pseudoscalars , which we leave for a future study .
we urge our experimental colleagues to look for our suggested signals at the lhc , and perhaps also reanalyze the existing collider data .
* acknowledgments : * we thank m. frigerio for useful discussions .
this work was supported by daad - dst ppp grant no .
d/08/04933 , and dst - daad project no .
int / daad / p-181/2008 .
g.b . acknowledges hospitality at t.u .
dortmund , ist - lisbon , lpt - orsay , and ictp - trieste , at different stages of this work .
was supported by dfg grant no .
pa 803/6 - 1 .
p.l . and h.p .
were also supported by the physics at the terascale helmholtz alliance working group : neutrino masses and lepton flavor violation at the lhc , and they acknowledge hospitality at the saha institute of nuclear physics , kolkata , during a part of this collaboration .
e. ma , new j. phys .
* 6 * , 104 ( 2004 ) [ arxiv : hep - ph/0405152 ] ; e. ma , arxiv:0705.0327 [ hep - ph ] ; g. altarelli and f. feruglio , rev .
phys . * 82 * , 2701 ( 2010 ) [ arxiv:1002.0211 [ hep - ph ] ] ; h. ishimori , t. kobayashi , h. ohki , y. shimizu , h. okada and m. tanimoto , prog .
suppl . * 183 * , 1 ( 2010 ) [ arxiv:1003.3552 [ hep - th ] ] .
e. ma , phys .
d * 61 * , 033012 ( 2000 ) [ arxiv : hep - ph/9909249 ] ; e. ma , phys .
d * 43 * , 2761 ( 1991 ) ; r. n. mohapatra , a. perez - lorenzana and c. a. de sousa pires , phys .
b * 474 * , 355 ( 2000 ) [ arxiv : hep - ph/9911395 ] ; p. f. harrison and w. g. scott , phys .
b * 557 * , 76 ( 2003 ) [ arxiv : hep - ph/0302025 ] ; w. grimus and l. lavoura , jhep * 0508 * , 013 ( 2005 ) [ arxiv : hep - ph/0504153 ] ; t. teshima , phys . rev .
d * 73 * , 045019 ( 2006 ) [ arxiv : hep - ph/0509094 ] ; r. n. mohapatra , s. nasri and h. b. yu , phys .
b * 639 * , 318 ( 2006 ) [ arxiv : hep - ph/0605020 ] ; s. morisi and m. picariello , int .
j. theor .
* 45 * , 1267 ( 2006 ) [ arxiv : hep - ph/0505113 ] .
r. n. mohapatra and g. senjanovic , phys .
* 44 * , 912 ( 1980 ) .
j. kubo , h. okada and f. sakamaki , phys .
d * 70 * , 036007 ( 2004 ) [ arxiv : hep - ph/0402089 ] .
s. l. glashow and s. weinberg , phys .
d * 15 * , 1958 ( 1977 ) .
z. z. xing , d. yang and s. zhou , phys .
b * 690 * , 304 ( 2010 ) [ arxiv:1004.4234 [ hep - ph ] ] .
f. j. botella , g. c. branco and m. n. rebelo , phys .
b * 687 * , 194 ( 2010 ) [ arxiv:0911.1753 [ hep - ph ] ] .
g. c. branco , w. grimus and l. lavoura , phys .
b * 380 * , 119 ( 1996 ) [ arxiv : hep - ph/9601383 ] .
a. s. joshipura and s. d. rindani , phys .
b * 260 * , 149 ( 1991 ) .
j. kubo , a. mondragon , m. mondragon and e. rodriguez - jauregui , prog .
phys . * 109 * , 795 ( 2003 ) [ erratum - ibid . * 114 * , 287 ( 2005 ) ] [ arxiv : hep - ph/0302196 ] .
y. yamanaka , h. sugawara and s. pakvasa , phys .
d * 25 * , 1895 ( 1982 ) [ erratum - ibid .
d * 29 * , 2135 ( 1984 ) ] .
d. lopez - val and j. sola , phys . rev .
d * 81 * , 033003 ( 2010 ) [ arxiv:0908.2898 [ hep - ph ] ] .
a. djouadi , j. kalinowski and m. spira , comput .
commun . * 108 * , 56 ( 1998 ) [ arxiv : hep - ph/9704448 ] .
s. schael _ et al . _
[ aleph collaboration and delphi collaboration and l3 collaboration and opal collaborations and lep working group for higgs boson searches ] , eur .
j. c * 47 * , 547 ( 2006 ) [ arxiv : hep - ex/0602042 ] .
s. schael _ et al . _
[ aleph collaboration ] , jhep * 1005 * , 049 ( 2010 ) [ arxiv:1003.0705 [ hep - ex ] ] .
s. chang , p. j. fox and n. weiner , jhep * 0608 * , 068 ( 2006 ) [ arxiv : hep - ph/0511250 ] .
s. chang , p. j. fox and n. weiner , phys .
lett . * 98 * , 111802 ( 2007 ) [ arxiv : hep - ph/0608310 ] . | discrete symmetries employed to explain flavor mixing and mass hierarchies can be associated with an enlarged scalar sector which might lead to exotic higgs decay modes . in this paper
, we explore such a possibility in a scenario with @xmath0 flavor symmetry which requires three scalar @xmath1 doublets .
the spectrum is fixed by minimizing the scalar potential , and we observe that the symmetry of the model leads to tantalizing higgs decay modes potentially observable at the cern large hadron collider ( lhc ) . `
pacs nos : 14.80.ec , 11.30.hv ` + ` key words : higgs boson , flavor symmetry ` sinp / tnp/2010/12 , th - do 10/10 * exotic higgs boson decay modes as a harbinger of @xmath0 flavor symmetry * + gautam bhattacharyya @xmath2 , philipp leser @xmath3 , and heinrich ps @xmath3 + + @xmath3 _ fakultt fr physik , technische universitt dortmund , 44221 dortmund , germany _ |
it is generally believed that type ib and type ic supernovae result from core collapse events of naked helium stars .
the helium stars are thought to be produced by the loss of the hydrogen envelope , via stellar winds mass loss from massive single stars or via mass transfers in close binary systems .
according to recent stellar models adopting the most up - to - date stellar winds mass loss rates @xcite , the final masses ( @xmath5 ) of helium stars produced by mass - losing single stars appear to be too massive to produce typical sne ib / c ( i.e. , @xmath6 @xmath0 at solar metallicity ) .
although the limiting mass for bh formation is not yet well determined , given their high binding energy , such massive progenitors of @xmath6 @xmath0 are likely to form black holes ( bhs ) , producing faint supernovae or no supernova at all ( cf . * ? ? ?
although very bright sne ib / c like sn 1998bw could be produced from such massive helium stars if , for example , powered by rapid rotation ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , such events are shown to be rare ( e.g. , * ? ? ?
* ; * ? ? ?
by contrast , helium stars with a wide range of masses ( @xmath7 ) can be made from @xmath8 @xmath0 primary components in close binary systems via the so - called case a / b mass transfer .
many of them may end their life as bright sne ib / c leaving neutron stars as remnants , if their final masses are less than about 7 10 @xmath0 .
population studies indeed show that close binary stars can produce a sufficient number of sne ibc to explain their observed rate , without the need of invoking single star progenitors ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
therefore , it is most likely that the majority of typical sne ibc are produced in binary systems .
the observational evidence for the connection between sne ibc and long gamma - ray bursts ( grbs ) has particularly motivated many observational studies to better understand sne ibc since the last decade ( see * ? ? ? * for a review ) .
theoretical stellar models of sne ibc progenitors are thus highly required nowadays .
the most comprehensive studies on the detailed characteristics of sne ibc progenitors in binary systems were conducted by @xcite ( hereafter , wlw95 ) using mass - losing pure helium star models , and by @xcite ( hereafter , wl99 ) using self - consistent binary star models .
although more recent theoretical studies on sne ibc progenitors in binary systems can be found in the literature , they have been focused on long grb progenitors or stellar populations , rather than on the detailed nature of typical sne ibc progenitors ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in this paper
, we revisit the problem of sne ibc progenitors in close binary systems using both binary star and single helium star models up to the neon burning stage , with updated physics of two important ingredients . one is rotation , which was not considered in wlw95 and wl99 , and the other is the mass loss rate of wolf - rayet ( wr ) stars ( sect . [ sect : review ] ) .
this paper is organized as follows . in sect .
[ sect : review ] , we briefly review recent developments of stellar evolution models regarding the effects of rotation and the wr star mass loss rate , arguing for the need of updated physics in binary star models . our adopted physical assumptions and numerical method are discussed in sect .
[ sect : method ] . in the following section ( sec . [ sect : rotation ] ) , using our binary star evolution models including the effect of rotation and the transport of angular momentum due to hydrodynamic instabilities and magnetic torques , we explore the role of tidal interaction and mass transfer in the redistribution of angular momentum in primary stars . in sec .
[ sect : sn ] , the nature of sne ibc progenitors is investigated in terms of final masses , masses of helium and hydrogen layers , radii and mass loss rates at the presupernova stage , assuming these properties do not significantly change from neon burning to core collapse . for this purpose
, we also present mass - losing single helium star models as a complement to our binary star models , given that the parameter space explored with our binary model sequences is limited .
we conclude the paper by discussing observational implications of our results , in sect .
[ sect : discussion ] .
rotation has particular roles in the evolution massive stars as it changes the stellar structure , induces chemical mixing and enhances mass loss due to stellar winds @xcite . during the last decade ,
several authors have calculated massive star models up to the pre - supernova stage , considering the redistribution of angular momentum and chemical species due to rotationally induced hydrodynamic instabilities , such as eddington - sweet circulations , and the shear instability @xcite .
although these models could explain some observational aspects such as surface abundances of cno elements of massive stars and wolf - rayet ( wr ) star populations at different metallicities ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , their adopted angular momentum transport mechanisms turned out to be too inefficient to explain the observed spin rates of stellar remnants @xcite .
i.e. , their models predict nearly two orders of magnitude higher spin rates of white dwarfs and young neutron stars than the observed ones . these models also imply that almost all wr stars can retain enough angular momentum to produce grbs , either by magnetar or collapsar formation depending on the final mass .
this gives a nearly 1000 times higher ratio of grbs to sne than the observationally implied value . on the other hand
, @xcite suggested that magnetic torques resulting from dynamo actions in differentially rotating radiative layers ( the so - called spruit - tayler dynamo ) should be the dominant angular momentum transport mechanism compared to the pure hydrodynamic instabilities .
recent magnetic models that adopt the spruit - tayler dynamo according to the prescription by @xcite are indeed more consistent with observations in terms of the spin rates of stellar remnants @xcite .
magnetic models also better explain the fact that long grbs are rare events compared to normal core collapse supernovae .
although the spruit - tayler dynamo mechanism is still subject to many uncertainties @xcite , these recent studies indicate that an efficient angular momentum transport mechanism comparable to what the spruit - tayler dynamo predicts is needed to understand the observations . the above discussion is based on single star models , and the role of rotation in the evolution of massive binary stars remained relatively unexplored .
massive stars in close binary systems are supposed to experience an exchange of mass and angular momentum via mass transfer and tidal interaction , and thus the evolution of binary stars is more complex than that of single stars . @xcite and @xcite presented , for the first time , self - consistent calculations of massive binary star evolution models including many relevant effects of rotation and binary interactions : the change of the stellar structure due to the centrifugal force , transport of angular momentum and chemical species due to rotationally induced hydrodynamic instabilities , tidal interaction , transfer of mass from the primary , accretion of mass and angular momentum of the secondary , and resulting changes of the orbit .
their work indicates that the secondary can be easily spun up by mass accretion even up to critical rotation , thus modulating the mass accretion efficiency by the interplay of the enhanced mass loss due to rotation from the spun - up secondary and the mass transfer from the primary .
this effect provides important clues to better understand the evolutionary paths of some observed x - ray and wr star binary systems as discussed by @xcite and @xcite .
their non - magnetic models also show that massive stars may end up with different core spin rates depending on the history of mass loss / gain during binary evolution , that could be related to the observational diversity of core - collapse supernovae .
more recently , @xcite and @xcite included the spruit - tayler dynamo in their binary models , and discussed possible evolutionary paths of massive binary stars towards long grbs . in this paper , we present new evolutionary calculations of magnetic ( i.e. , the spruit - tayler dynamo is included ) massive binary stars for a large parameter space , focusing on the evolution of the primary stars to investigate the nature of typical sne ibc progenitors . wlw95 employed the mass - dependent mass loss rate of wr stars of @xcite for their helium star models , and wl99 used those of @xcite and @xcite .
both studies concluded that the final masses of sne ibc progenitors in binary systems should converge to a limited range of 2.2 3.6 @xmath0 , even for an initial helium star mass of 20 @xmath0 .
later developments of the wr wind theory considering clumpies pointed out , however , that the wr mass loss rates used by wlw95 and wl99 are significantly overestimated @xcite . for example , fig .
[ fig : wrwind ] shows that , for helium stars on the zero age main sequence , the mass loss rate given by @xcite is almost an order - of - magnitude lower than that of hamann et al .
used in wl99 .
the most recent theoretical models of wr winds by @xcite and @xcite also give wr mass loss rates compatible to the nugis & lamers rate ( cf .
figure 1 in * ? ? ?
the reduced wr mass loss rates have been considered in many recent massive single star models ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , suggesting large final masses of wr stars at the presupernova stage ( @xmath9 @xmath0 at solar metallicity ) compared to @xmath10 @xmath0 found by @xcite .
as argued in the introduction , this leads to the interesting conclusion that most core collapse events occurring in single wr stars should not produce bright supernovae , rendering the binary star channel even more important for sne ibc .
surprisingly , the effect of the new wr mass loss rate on sne ibc progenitors in close binary systems has not been much discussed or largely overlooked . to our knowledge , @xcite are the only authors who addressed this issue in detail .
although wl99 also presented some model sequences with a reduced wr mass loss rate , their adopted reduction factor was rather modest ( 2 times smaller than the hamann et al .
rate ) . here
we consider the hamman et al .
wr mass loss rate reduced by factors of 5 and 10 , which reflects the most recent result as discussed above .
we have used the same stellar evolution code as in @xcite , that follows the simultaneous evolution of the two stellar components of a binary system .
the effect of the centrifugal force on stellar structure is considered following @xcite .
the transport of angular momentum and chemical elements is treated as diffusion considering eddington sweet circulations , the shear instability , the goldreich - schubert - fricke instability and the spruit - tayler dynamo , as described in @xcite .
we use @xmath11 for the semi - convection parameter @xcite , as in @xcite .
the stellar winds mass loss rate is determined following @xcite with a metallicity scaling of @xmath12 @xcite for the main sequence phase .
wr wind mass loss rates are computed according to hamman et al . with a correction factor @xmath13 , and with a metallicity dependence of @xmath14 @xcite : @xmath15 we use @xmath16 or 10 in most model sequences , which means the wr mass loss rate by hamman et al .
is lowered 5 or 10 times , to consider the most recent estimates ( see fig .
[ fig : wrwind ] ) .
the enhancement of stellar winds mass loss due to the centrifugal force is considered by @xmath17 ~~,\ ] ] where @xmath18 with @xmath19 and @xmath20 is the eddington factor @xcite . here , to prevent a singularity that may occur as @xmath21 approaches @xmath22
, the mass loss rate is limited to @xmath23 , where @xmath24 is the thermal time scale of the star .
the binary orbit is assumed to be circular , and the roche lobe radius is determined according to the approximation of @xcite .
the mass loss rate of the roche - lobe filling component through the first lagrangian point is implicitly computed using the method given by @xcite .
the equation of motion of a test particle is numerically solved to calculate the amount of angular momentum of the accreted matter if the transfered matter directly hits the secondary star , and the keplerian value is assumed otherwise @xcite .
the change of the orbital period due to mass transfer and stellar wind mass loss is considered according to @xcite .
we follow @xcite to determine the amount of the specific angular momentum carried away from the orbit by stellar winds .
tidal synchronization is considered following @xcite ( see also * ? ? ?
we assume a synchronization time scale according to @xcite who considered tidally driven meridional circulations as the main mechanism for tidal dissipation : @xmath25 where @xmath26 denotes the mass ratio and @xmath27 the orbital separation .
this prescription gives a much shorter time scale than that given by @xcite .
given that the physics of tidal dissipation is much debated in the literature @xcite , we introduce a parameter @xmath28 to investigate how an extremely fast / slow synchronization may influence the results . in most cases , however , we use @xmath29 . a few sequences are also computed with @xmath30 of @xcite for comparison : @xmath31 where @xmath32 is the moment of the star , and @xmath33 a constant measuring the coupling between the tidal potential and the gravity mode . using the data of table 1 in @xcite
, we constructed a fitting formula for @xmath33 as the following : @xmath34 where @xmath35 is the radius of the convective core .
note that both prescriptions by tassoul and zahn are not appropriate for a star with a convective envelope .
however , the role of tidal synchronization is significant only on the main sequence , and not important in late evolutionary stages as discussed below .
we computed 45 model sequences for initial masses of the primary star mostly from 12 to 25 @xmath36 at two different metallicities ( @xmath37 0.02 and 0.004 ) , for different mass ratios , initial orbital periods , and wr mass loss rates , as summarized in table .
the initial rotational velocity at the equatorial surface of each star is set to be 20% of the keplerivan value .
we could not calculate more massive systems because of a numerical difficulty encountered during the mass transfer phases , except for seq .
26 where a primary star of 60 @xmath0 is considered with a rather large wr mass loss rate ( i.e. , @xmath38 ) .
the adopted initial orbital periods corresponds either to case a or to case b mass transfer . in the present study , we do not consider case c systems , but briefly discuss the possible outcomes of case c mass transfer in sect .
[ sect : dischydrogen ] .
the evolution of the primary stars is followed up to neon burning in most cases .
we also present non - rotating single helium star models to discuss sne ibc progenitors in binary systems with initial masses larger than 25 @xmath0 , and also to compare them with binary star models ( sect .
[ sect : sn ] ) .
in this section , we focus our discussion on the evolution of primary stars and investigate whether binary evolution via case a or case b mass transfer could lead to diverse pre - collapse conditions of sne ibc in terms of the amount of core angular momentum . although the evolution of mass - accreting secondary stars is a matter of extreme interest as discussed in @xcite , @xcite and @xcite , it is beyond the scope of this paper . here , we first present some results including the spruit - tayler dynamo with our fiducial assumption on synchronization time ( i.e. , @xmath39 ) , showing that the final amount of angular momentum in the core of the primary star is not much affected by different histories of mass loss ( i.e. , case ab or case b ; sect .
[ sect : fiducial ] ) .
then , we discuss the influences of different assumptions on tidal synchronization and transport process of angular momentum ( sect .
[ sect : nonfiducial ] ) .
the evolution of the primary star in a close binary system is characterized by the rapid loss of mass due to roche - lobe overflow . as an example , the evolution of the primary star in seq .
14 is described in figs .
[ fig : kippseq9 ] and [ fig : hrseq9 ] , where our fiducial value of @xmath39 is adopted , including the spruit - tayler dyanamo .
the binary system initially consists of a @xmath40 star and a @xmath41 star in a 4 day orbit .
mass transfer starts at @xmath42 , when the helium mass fraction in the hydrogen burning core has increased to 0.94 .
the mass transfer rate rises up to @xmath43 , which roughly corresponds to @xmath44 where @xmath45 and @xmath46 denote the mass and the kelvin - helmoltz time scale of the primary star , respectively .
the primary mass decreases to @xmath47 by the end of the case a transfer ( see fig .
[ fig : chem ] ) .
the second roche - lobe overflow begins at @xmath48 when the envelope of the primary star expands due to hydrogen shell burning during the helium core contraction phase ( case ab mass transfer ) .
the primary star loses most of the hydrogen envelope as a result , exposing its helium core of @xmath49 having a small amount of hydrogen ( @xmath50 @xmath0 ) in the outermost layers , as shown in the third panel of fig .
[ fig : chem ] .
although the star remains compact ( @xmath51 ) during core helium burning , helium shell burning activated after core helium exhaustion leads to the expansion of the envelope up to @xmath52 ( see fig .
[ fig : hrseq9 ] ) during core carbon burning .
a case abb mass transfer does not occur , however , due to the large orbital separation ( @xmath53 ) at this stage , while it does occur in many other sequences .
the final mass at the end of the calculation ( neon burning ) is @xmath54 .
the mass of hydrogen decreases to 0.0015 @xmath0 at the end , and the remaning mass of helium is 1.49 @xmath0 , as shown in the last panel of fig .
[ fig : chem ] . the star is likely to eventually explode as a type ib supernova given the rather thick helium envelope with a very thin hydrogen layer , but it might also appear as type iib if the supernova were found within several days after the explosion ( see sect . [
sect : hydrogen ] ) .
the lower panel of fig .
[ fig : hrseq9 ] shows that the mass transfer is not conservative .
when the secondary star reaches critical rotation as a result of the accretion of angular momentum , the stellar wind mass loss rate increases so drastically as to prevent efficient mass accumulation ( see * ? ? ?
* for a more detailed discussion on this effect ) .
the ratio of the accreted mass in the secondary star to the transferred mass from the primary star is about 0.83 during the case a mass transfer , and 0.41 during the case b mass transfer .
the previous calculations by wl99 and @xcite show that the non - conservative mass transfer leads to a shorter orbit than in the case of conservative mass transfer , in general .
this effect should be kept in mind in the following discussion on the evolution of the binary orbit and its consequences .
the evolution of rotation in the primary star is shown in figs .
[ fig : totjsync ] , [ fig : omegasync ] and [ fig : jcoreseq9 ] . as shown in fig .
[ fig : totjsync ] , the total angular momentum of the primary star rapidly decreases in the beginning as a result of tidal interaction , until the star is completely synchronized with the orbit when @xmath55 .
[ fig : omegasync ] shows the evolution of the angular velocity inside the primary star from zams until it reaches complete synchronization . note that the star keeps rotating almost rigidly , even though the tidal interaction causes redistribution of angular momentum from the outermost layers of the star @xcite .
this results from the very short time scale of the transport of angular momentum in the star ( about 300 years ) due mainly to convection and the spruit - tayler dynamo , in the convective and radiative layers respectively . later on ,
the angular velocity of the primary star still remains coupled with the orbital motion , until decoupling starts during case b mass transfer ( see below ) .
interestingly the total angular momentum of the primary star gradually increases from the initial synchronization until the onset of case a mass transfer ( fig .
[ fig : totjsync ] ) .
this is because the star significantly expands , while the change of the orbital separation remains small durning this period , as shown in the third and last panels in fig .
[ fig : jcoreseq9 ] .
however , it rapidly decreases again during the case a and ab mass transfer phases , as explained below .
[ fig : jcoreseq9 ] shows that angular momentum in the core of the primary star is mostly removed during the mass transfer phases ( case a and case ab ) .
it should be noted , however , that the mechanism for braking the core is different for each case . during the case
a mass transfer , the synchronization time scale according to eq .
( [ eq3 ] ) remains very short ( @xmath56 ) compared to the mass transfer time scale ( @xmath57 ) .
the decrease of the core angular momentum during the case a mass transfer results from the synchronization that occurs even when the orbit is rapidly widened due to mass exchange ( fig .
[ fig : jcoreseq9 ] ) . during the case ab mass transfer , the further increase of the orbital separation significantly weakens the role of synchronization for the redistribution of angular momentum .
however , both the spruit - tayler dynamo and mass loss lead to rapid braking of the thermally contracting helium core .
further significant core - braking by the spruit - tayler dynamo occurs during the co core contraction phase , and the mean specific angular momentum of the innermost @xmath58 becomes about @xmath59 at the neon burning phase . in seq .
20 , the initial masses of the stellar components are the same as in seq . 14 but the initial period is large enough that the first mass transfer occurs during the helium core contraction phase ( case b mass transfer ) . avoiding case a mass transfer , the primary star retains more angular momentum in the core at the end of main sequence than in seq . 14
however , the core loses more angular momentum during the helium core contraction phase than in seq .
14 , as stronger magnetic torques are exerted , mainly due to the more massive envelope ( fig .
[ fig : caseb ] ) . at neon burning
, @xmath60 is obtained , which is similar to that in seq .
14 .
the effect of synchronization is negligible in seq .
15 , where @xmath61 is adopted . as shown in fig .
[ fig : jcoreseq10 ] , a rather rapid decrease of @xmath62 occurs during and after the case a mass transfer , which is a combined effect of mass loss and magnetic torques : mass loss carries away angular momentum from the envelope , and magnetic torques brake the core rotation subsequently .
the core is further slowed down during the helium core and co core contraction phases , resulting in @xmath63 at the neon burning phase .
note that this value is very close to that in seq .
we find that the result with zahn s prescription for synchronization ( seq .
17 ) is not much different from that of seq .
15 where synchronization is negligible .
the synchronization time scale according to zahn is sensitive to the ratio of the convective core size to the stellar radius ( @xmath64 , see eqs .
( [ eq4 ] ) and ( [ eq5 ] ) ) .
this ratio continuously decreases as the star evolves , and thus @xmath30 continuously increases to such an extent that the effect of synchronization can be ignored when case a mass transfer starts .
when @xmath65 with tassoul s prescription is used ( seq .
16 ) , on the other hand , synchronization becomes important even after the case ab mass transfer phase , and the primary is further spun down by tidal interaction , giving @xmath66 at the neon burning phase . in seq .
18 where the spruit - tayler dynamo is not included , the core is spun down due to synchronization during the case a mass transfer phase ( fig .
[ fig : jcoreseq10 ] ) as in the corresponding magnetic case ( seq .
14 ; fig .
[ fig : jcoreseq9 ] ) .
the spin - orbit coupling becomes significantly weakened as the orbit widens after the case a mass transfer phase . despite a significant amount of mass is lost via case ab mass transfer , the core in the primary star retains most of the remaining angular momentum in the following evolutionary stages .
this is because the chemical gradient across the boundary between the helium core and the hydrogen envelope effectively prohibits the transport of angular momentum ( cf . * ? ? ?
* ; * ? ? ?
the core angular momentum at neon burning is thus about 10 times larger ( @xmath67 ) than in the corresponding magnetic case . as shown in the above examples , in the model sequences with the spruit - tayler dynamo , all of the primary stars retain similar amounts of angular momentum ( a few @xmath68 ) in the innermost @xmath58 at neon burning regardless of the detailed history of mass transfer , unless synchronization is extremely fast as in seq .
16 ( see table [ tab1 ] ) .
according to the spruit - tayler dynamo , magnetic torques exerted to the core become stronger with a higher spin rate , a larger degree of differential rotation between the core and the envelope , and a heavier radiative envelope .
therefore , although winds or roche - lobe overflows reduce the size of the hydrogen envelope and remove angular momentum from the star , this in turn weakens the torque exerted to the core , and vice versa .
the remarkable convergence of @xmath69 to a few @xmath68 in our model sequences , even for different wind parameters and metallicities as shown in table [ tab1 ] , can be explained by this self - regulating nature of the spruit - tayler dynamo .
this result indicates that not much diversity is expected in sne ibc progenitors produced via case a or case ab / b mass transfer , in terms of rotation : most of sne ibc of a similar progenitor mass may leave neutron stars with a similar spin rate .
however , other types of binary interactions still may lead to various final rotation periods in sn progenitors ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* for such examples ) .
in the literature , the detailed history of binary interactions such as mass transfer and tidal interaction is often neglected , and only the evolution of pure helium stars to discuss sne ibc progenitors in binary systems is investigated ( e.g. wlw95 ; @xcite ) .
although this approach is very useful , our binary star models indicate that complications resulting from binary interactions have important consequences in sn ibc progenitors . to address this issue , here we also present evolutionary models of non - rotating single helium stars with initial masses of 2.8 20 @xmath0 at @xmath70 for comparison with binary star models , as presented in table [ tab2 ] .
this also enables us to study sn ibc progenitors with initial masses higher than 25 @xmath0 , which are lacking in our binary model sequences .
these helium star models were calculated up to neon burning , with the wr mass loss rate given in eq .
( [ eq1 ] ) with @xmath71 5 or 10 , as in the binary models . in the following
, we focus our discussion mostly on the results with @xmath16 , while the influence of the wind parameter is briefly discussed . the final mass of the primary stars in a close binary systems is largely determined by the winds mass loss during core helium burning , if the initial mass ( @xmath72 ; i.e. , the mass right after case ab or case b mass transfer ) is significantly larger than about 3.0 @xmath0 .
however , for a less massive helium star , the expansion of the helium envelope becomes so dramatic during co core contraction , and/or during core carbon burning , that it can lead to another mass transfer phase : case abb or case bb ( e.g. , compare the final radius of the 2.8 @xmath0 helium star model with those of more massive helium star models in table [ tab2 ] ) .
the impact of this mass transfer phase becomes more important for a less massive helium star . in seqs . 1 , 2 and 4 where @xmath73 2.1 2.3 @xmath0 , the rapid loss of mass during case abb / bb transfer reduces the total mass of the primary stars to such an extent that they may not explode as supernovae , but die as white dwarfs .
in seq . 3 , where @xmath74 @xmath0 , the carbon - oxygen core can grow beyond the chandrasekhar limit despite the significant loss of mass ( i.e. , about 1 @xmath0 ) during the case bb phase .
the remaining helium mass in the envelop is only about 0.18 @xmath0 at core carbon exhaustion . for the other sequences at solar metallicity ( seqs . 5 , 8 , 9 , 10 , 11 , 13 , 21 ) where @xmath75 , the amounts of mass loss via case abb / bb transfer until core neon burning are rather moderate , varying from 0.02 to 0.1 @xmath0 .
the wr wind mass loss rate has significant consequences in systems where @xmath76 3.0 @xmath0 .
the previous study by wlw95 concluded that the final masses of single helium stars with initial masses of 3 20 @xmath0 should converge to @xmath77 2.2 3.5 @xmath0 .
as expected from the lower wr mass loss rate adopted in this study , our helium star models give a significantly wider range of final masses ( @xmath5 ) at the given metallicity of 0.02 : @xmath78 2.9 7 @xmath0 ( 2.9 10 @xmath0 ) from helium stars of 3 20 @xmath0 for @xmath16 ( @xmath79 ; table [ tab2 ] ; cf . fig .
[ fig : mimf ] ) .
our binary star models also give generally larger final masses for sn ibc progenitors than those in wl99 : @xmath80 1.64 4.5 @xmath0 for @xmath81 from @xmath8213 25 @xmath0 , compared to @xmath80 2.0 3.4 @xmath0 in wl99 is determined by the detailed history of case abb / bb transfer , which depends on uncertain parameters such as the semi - convection efficiency . ] .
figure [ fig : dmhe ] shows the amount of helium ( @xmath83 ) as a function of the final mass in the sn ibc progenitor models with @xmath84 at solar metallicity . in general , both binary and helium star models predict large amounts of helium in the envelope
compared to the results of wlw95 and wl99 except for relatively low mass systems of @xmath85 @xmath0 .
note that only rather massive progenitors with @xmath86 @xmath0 can be significantly helium deficient ( i.e. , @xmath87 ) . for the systems with @xmath88 @xmath0 ,
the role of case abb / bb mass transfer becomes important .
in particular , the small amount of helium ( @xmath89 ) in seq .
3 shows that helium deficiency can also be achieved by the so - called case bb / abb mass transfer from relatively low mass helium cores ( @xmath90 ) .
therefore , we expect two different classes for type ic progenitors at solar metallicity : one with @xmath91 @xmath0 and the other with @xmath92 5.5 @xmath0 ( see also * ? ? ?
* ; * ? ? ?
obviously , the upper limit of @xmath5 for potential type ic progenitors increases with a smaller mass loss rate : with @xmath93 , we have @xmath92 8.9 @xmath0 . in our binary star models , case ab or case b mass transfer does not completely remove hydrogen from the primary stars , as shown in fig . [
fig : chem ] .
the resulting wr mass loss is therefore weaker than that from the corresponding pure helium star .
this explains the fact that for a given wr mass loss rate , the binary star models with @xmath94 @xmath0 predict somewhat larger @xmath83 than the single helium star models do as shown in fig . [ fig : dmhe ] . more importantly , the binary star models show that , even at @xmath95 , small amounts of hydrogen can be retained up to the pre - supernova stage for a certain range of the final mass .
[ fig : dmh ] ( see also table [ tab1 ] and fig . [
fig : chem ] ) shows the total mass of hydrogen in the primary stars at neon / oxygen burning phase . the time span from this stage to core collapse
is supposed to be less than about 10 yr . the wr winds mass loss rate from the primary star models shown in the figure is about @xmath96 .
some of the primary stars in this sample are still undergoing case abb or bb mass transfer ( seqs . 5 , 8 , 9 , 13 , & 21 and many of the smc models ; see table [ tab1 ] ) , but the mass transfer rate is only about @xmath97 .
therefore , the amount of hydrogen at core collapse should remain close to the values given in the figure .
fig [ fig : dmh ] indicates that the presence of hydrogen is only expected for @xmath98 \la 3.7 $ ] , at solar metallicity with @xmath99 .
this results from two different reasons for different mass ranges . for @xmath100 ,
primary stars expand during the carbon burning phase to much larger radii than more massive ones do .
the resulting mass loss rates via case abb or bb transfer thus become large enough to completely remove hydrogen from the primary stars by the time of core collapse . for @xmath101 , on the other hand , wr winds are rather strong and can remove hydrogen from the primary stars during the core helium burning phase .
the upper limit of @xmath5 for the presence of hydrogen thus depends on the adopted mass loss rate .
as shown in the figure , with @xmath93 , it increases to about 4.5 @xmath0 at solar metallicity . at smc metallicity ,
the stellar wind effect is less significant and even rather massive supernova progenitors of about 7 @xmath0 can retain fairly large amounts of hydrogen ( @xmath102 ) .
we discuss implications of this result for supernova types in sect .
[ sect : dischydrogen ] .
relatively low mass helium stars usually experience a rapid expansion of the envelope during the core carbon burning phase .
interestingly , our new models predict much larger radii at the pre - supernova stages than what the wlw95 models do , as shown in fig .
[ fig : radius ] .
the first reason for this difference is the updated opal opacity table by @xcite , which causes a strong iron bump at around @xmath103 .
another reason is the smaller wr mass loss rate adopted in the present study , and the resultant thicker helium envelope in our models .
the radius is further affected by the complication of binary interaction : the presence of a thin hydrogen layer in some of the binary models results in a more extended envelope than in the corresponding single helium star models .
it is also worth noting that the radii of the binary star models at @xmath104 are smaller than those of the single helium star models .
this is because the primary stars in these sequences are filling the roche - lobe , and the envelope can not expand beyond it .
the metallicity effect is somewhat subtle .
the hydrogen and helium layers in sn ibc progenitors become thicker due to the reduced mass loss rate for lower metallicity , while the opacity due to metals becomes smaller .
as these two effects compensate each other , the radii of our sn ibc progenitor models at both smc and solar metallicities are found to be similar .
finally , it should be noted that the radius depends on the final mass . in general , a less massive progenitor tends to have a larger radius at the presupernova stage .
this might make mixing of nickel into the helium rich layer induced by the rayleigh - taylor instability during the supernova explosion more efficient for a less massive sn ibc progenitor , as implied by the simulations of @xcite and @xcite , resulting in more prominent helium lines in the spectra ( see sect .
[ sect : discussion ] below ) .
[ fig : mdot ] shows the mass loss rate due to wr winds at the central neon / oxygen burning phase from different sn ibc progenitor models .
the mass loss rate ranges from @xmath105 to @xmath106 , depending on the final mass , at solar metallicity .
the corresponding wind velocity ( i.e. , the escape velocity ) changes from @xmath107 to @xmath108 . the circumstellar interaction of such winds from primary stars in binary systems
are supposed to be very complex because of the orbital motion and the wind - wind interaction with the secondary star .
some of the primary star models shown in the figure are still undergoing case abb or bb mass transfer , which may add to the complexity .
therefore , the nature of the circumstellar materials around sn ibc progenitors in a binary system may not follow the simple relation of @xmath109 that is expected for single wr star progenitors .
we have presented new evolutionary models of massive close binary stars , considering tidal interaction , and transport of angular momentum and chemical species due to rotationally induced hydrodynamic instabilities and the spruit - tayler dynamo .
we have investigated the redistribution of angular momentum in the primary star .
although mass transfer and tidal interaction can significantly affect the evolution of the rotation velocity of the primary star on the main sequence , the amount of angular momentum retained in the core in the late evolutionary stages is rather insensitive to the previous history of such binary interactions because of the self - regulating nature of the spruit - tayler dynamo .
we have also calculated non - rotating , mass - losing single helium star models and compared them with our primary star models in binary systems .
our models adopt a much lower wr mass loss rate than in the previous studies by wlw95 and wl99 , and predict some new important properties of sn ibc progenitors accordingly .
the following discussions are based on the models with @xmath16 , unless otherwise specified . 1 .
the final masses of sn ibc progenitors in binary systems at @xmath110 @xmath111 are not limited to @xmath112 as predicted by wlw95 and wl99 , but a more wide range @xmath5 is expected ( i.e. , @xmath113 from @xmath114 @xmath0 ; see fig . [
fig : mimf ] ) .
2 . at @xmath115 @xmath111 , significant deficiency of helium (
@xmath116 @xmath0 ) is expected for @xmath117 and for @xmath118 ( fig . [
fig : dmhe ] ) .
rather large amounts of helium up to 1.5 @xmath0 are expected for the other final mass range ( i.e. , @xmath119 5.5 @xmath0 ) . at @xmath115 @xmath120 ,
no such helium deficient sn progenitors are expected for the considered initial masses ( @xmath121 @xmath0 ) .
a thin layer of hydrogen with @xmath122 @xmath0 is predicted for sn ibc progenitors with @xmath123 at @xmath124@xmath111 , and @xmath125 at @xmath115@xmath120 , respectively ( fig .
[ fig : dmh ] ; table [ tab1 ] ) .
4 . most sn ibc progenitors with @xmath126 @xmath0 rapidly expand during core carbon burning , resulting in @xmath127 at the presuprnova stage .
this is much larger than found in wlw95 ( fig .
[ fig : radius ] ) .
compact progenitors of @xmath128 are only expected for a relatively high mass ( @xmath129 @xmath0 at @xmath110 @xmath111 and @xmath130 @xmath0 at @xmath110 @xmath120 ; fig .
[ fig : radius ] ) .
our binary star models show that the mass transfer during helium core contraction ( case ab or case b ) in a close massive binary system can not remove the hydrogen envelope promptly enough to avoid the core braking due to the spruit - tayler dynamo during the helium core contraction phase .
comparison of our binary star models with the single star models by @xcite and @xcite indicate that the amount of angular momentum retained in the core of the primary star at the presupernova stage should not be much different from those found in single star models if the spruit - tayler dynamo is adopted .
i.e. , a specific angular momentum of a few @xmath68 in the innermost @xmath771.4 @xmath0 at the presupernova stage is expected in both single and binary stars .
this value is smaller by one or two orders of magnitude than what is necessary to make a long gamma - ray bursts by magnetar or collapsar formation , or very energetic supernovae ( hypernovae ) powered by rapid rotation and strong magnetic fields ( e.g. * ? ? ?
* ) , although it may suffice to produce millisecond pulsars @xcite . together with the work by @xcite , our results thus indicate that binary interactions with case ab / b mass transfers at @xmath131 may not particularly enhance the production of strongly rotation - powered events like long grbs or hypernovae .
this is consistent with the observational evidence that such events are rare compared to normal core collapse events ( e.g. , * ? ? ?
* ; * ? ? ?
this also confirms the theoretical consensus that other evolutionary paths are needed to produce long grbs associated with sn ibc , such as the quasi - chemically homogeneous evolution of a metal poor star @xcite , tidal spin - up of a wr star in a very close binary system with a neutron star or black hole companion ( e.g. * ? ? ?
* ; * ? ? ?
* ) , or binary evolution with case c mass transfer with some specific conditions @xcite .
larger radii of our sn ibc progenitor models than those previously found should have consequences in shock break - outs and bolometric light curves .
for instance , a shock break - out from a larger envelope would be marked by a lower photosperic temperature .
detailed comparison of numerical calculations with observational data may thus give strong constraints on sne ibc progenitor properties ( e.g. , * ? ? ? * ) .
recent discovery of the x - ray outburst with sn 2008d by @xcite indeed suggests the usefulness of such a study for the probe of supernova progenitors ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , for which our new models would provide ideal input .
we will address this issue in a forthcoming paper .
although the weak signature or no evidence of helium in sne ic spectra may indicate the deficiency of helium in their progenitors , it is not well known how much helium can be hidden in the supernova spectra .
it may also depend on the degree of mixing of nickel into helium rich layers @xcite .
if we assume 0.5 @xmath0 , for instance , as the maximum amount of helium allowed for hiding helium lines in sn spectra , our models indicate that most sne ic progenitors at solar metallicity should belong to two distinct classes in terms of both zams and final masses , as summarized in fig .
[ fig : mimf ] for the binary systems that undergo case b mass transfer .
the same conclusion was also drawn by wl99 and @xcite .
but pols & dewi considered different types of binary systems ( see below ) , and the finding of `` two mass classes '' for the final masses was not obvious in wl99 due to the very high wr rate adopted in their study , although it was clearly seen for the zams masses .
if we assume that primary stars of @xmath132 produce low - mass - class sne ic via case bb mass transfer , about 62% of the sne ic from case b systems should belong to the high mass class and the rest ( @xmath7738% ) to the low mass class , at @xmath133 ( see fig .
[ fig : mimf ] ) .
it is important to note that the two classes are produced by different mechanisms .
the high mass class of sne ic progenitors ( i.e. , @xmath134 @xmath0 in fig [ fig : mimf ] ) is a consequence of wr winds mass loss , while the low mass class results from case abb / bb mass transfer as discussed in sect . [ sect : masses ] .
the zams mass range for the low mass class may not be much affected by metallicity , while it should be widened with increasing metallicity for the high mass class .
this leads to the conclusion that the low and high mass classes would dominate at low and high metallicities respectively .
it should also be noted that the final mass range for the low mass class may not change much for different metallicities , while it may decrease with increasing metallicity for the high mass class , due to the increasing wr winds mass loss rates , as implied by the result of wl99 . in the massive close binary systems considered in this paper ,
the primary star masses become much smaller than those of the secondary stars when case abb or case bb mass transfer begins .
if the companion star mass were lower than the helium star in a close binary system , the mass transfer rate should become higher than in the systems of the present study .
for example , if a helium star is located in a very short period binary system ( @xmath135 ) with a less massive companion ( e.g. a neutron star ) , mass transfer from the helium star may occur rapidly enough to make a helium - deficient carbon star , even for @xmath136 6.0 @xmath0 as shown by @xcite , @xcite and @xcite .
the final masses of such sn ic progenitors may range from 1.5 @xmath0 to 3.0 @xmath0 .
this scenario was also suggested by @xcite to explain the fast light curve of type ic sn 1994i .
the zams mass of such a sn ic should be in the range of 12 20 @xmath0 @xcite .
however , such close helium star plus neutron star systems are supposed to rarely form , and might not contribute much to the population of sne ic , compared to the systems considered in this study .
our results should have several observational consequences .
as mentioned above , the population of sne ic should be dominated by the high mass class for @xmath137 .
they would have higher zams and final masses than those of typical sne ib progenitors ( cf . * ? ? ?
* ; * ? ? ?
given that the parameter space for the high mass class sne ic may become larger with a higher wr mass loss rate , the number ratio of sne ic to sne ib , and that of high mass class sne ic to low mass class sne ic should increase with increasing metallicity ( cf .
* ; * ? ? ?
* ; * ? ? ?
the existence of the two mass classes of sne ic progenitors may be related to some aspects of the observational diversity of sne ic .
for example , sne ic of the low mass class is likely to be characterized by rather fast declining light curves and low luminosities ( cf . * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , implying that the observed population of sne ic is likely to be baised to high - mass - class sne ic .
we should also note the huge difference of the binding energy between the two classes .
the binding energy of the envelope above 1.4 @xmath0 in the sn ic progenitor star model of seq . 3 ( @xmath138 @xmath0 ) is only about @xmath139 erg , while it should be one or two orders of magnitude higher for a sn ic progenitor with @xmath140 @xmath0 ( see table [ tab2 ] ) . as a consequence
, the energetics of sne ic might be systematically different for the two different classes .
on the other hand , the assumption of @xmath141 @xmath0 for sn ic progenitors leads to a ratio of type ic to ib supernova rate ( ic / ib ratio ) of about 0.4 from binary systems at solar metallicity .
have @xmath142 @xmath0 with @xmath143 @xmath0 ( see table [ tab1 ] ) .
this implies that the low - mass class sne ic would predominantly occur in binary systems at this metallicity .
if we assume stars with @xmath144 would produce low - mass class sne ic as in the case of solar metallicity , the sn ic / sn ib ratio would be about 0.1 at smc metallicity .
the exploration of the exact mass range for the low - mass class sne ic is a time - consuming task , and we plan to investigate this in near future . ]
this appears in contradiction with recent observations that indicate rather a high ic / ib ratio of about 2.0 ( e.g. * ? ? ?
this discrepancy would become even larger with @xmath145 , as implied by fig .
[ fig : mimf2 ] .
this raises a question on the nature of sn ic progenitors , and it should be kept in mind that we still do not fully understand what distinguishes sn ic progenitors from those of sn ib .
a recent work by @xcite indicates that the mass fraction of helium in the outermost layers ( @xmath146 ) , rather than the total mass of helium , may be more relevant for the presence of helium lines in supernova spectra .
specifically , it is shown that if helium is well mixed with co material such that @xmath146 becomes less than about 0.5 , helium lines are not seen in early time spectra , despite rather a large total amount of helium ( @xmath147 @xmath0 ) , if non - thermal excitation is absent . in our progenitor models ,
such a small @xmath146 is realized only for @xmath92 5.5 @xmath0 at solar metallicity ( with @xmath16 ) .
this is not different from the above - discussed mass limit for having @xmath2 @xmath0 , implying that the initial mass range for sn ic progenitors would not change much even if we adopted @xmath146 as a criterion , at least for the high mass class . on the other hand
, we have @xmath148 in the primary star of seq . 3 at carbon exhaustion while the total mass of helium is less than 0.2 @xmath0 .
this imples that the initial mass range for the low mass class sn ic progenitors might be affected if the condition of @xmath149 for sn ic progenitors were applied .
but @xcite did not yet calculate such a low mass sn progenitor model ( @xmath91 @xmath0 ) , and their analyses were limited to early times of supernovae .
it remains to be an important subject of future work to systematically investigate which types of supernova progenitors would lead to the presence or absence of helium lines in the supernova spectra at different epochs , including the effect of non - thermal excitation .
therefore , the above discussion on type ic progenitors based on the total amount of helium should only be considered indicative at this stage .
it is interesting that , at @xmath133 , the presence of a thin hydrogen layer is only expected for a limited range of the initial / final mass of sn ib progenitors , as shown in figs [ fig : dmh ] , and [ fig : mimf ] .
the detection of hydrogen absorption lines at high velocity has been indeed reported in many sne ib ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , in favor of our model prediction for the presence of a thin hydrogen layer in sne ib progenitors .
this might provide a strong constraint for the progenitor masses of observed sne ib , in principle .
note also that explosions of such helium stars with thin hydyrogen layers could be recognized as sn iib rather than ib , if hydrogen lines were detected short after supernova explosion , e.g. , as in the case of sn 2008ax @xcite and as recently discussed by @xcite and @xcite .
the radii of these progenitor models range from @xmath150 to @xmath151 .
they may corredpond to the compact category of sn iib progenitors , which is discussed in @xcite .
the relatively low ejecta masses of such sne iib are consitent with our model predictions . on the other hand , case c mass transfer can also leave helium cores covered with small amounts of hydrogen envelope . as the life time of such stars made via case
c mass transfer should be rather short , they can retain much more hydrogen ( @xmath152 @xmath0 ) , than what is predicted from our binary models with case ab / b mass transfer .
such a star may eventually explode as a sn iib like sn 1993j ( e.g. * ? ? ?
* ; * ? ? ?
* ) , with a much extended envelope ( @xmath153 ) .
therefore , the two categories of sn iib progenitors according to their sizes , which has been recently suggested by @xcite , may be understood within the framework of binary evolution ; sne iib of the compact type may be produced via case ab / b mass transfer ( especially at @xmath154 ) , and sne iib of the extended type via case c mass transfer .
this work is supported by the doe scidac program ( doe de - fc02 - 06er41438 ) , the nsf grant ( nsf - arra ast-0909129 ) , and the nasa theory program ( nnx09ak36 g ) .
we are grateful to luc dessart for useful discussions .
cccccccccccccc 1 & 0.02 & 12 & 8 & 3.0 & 5 & b+bb & 57.9 & 1.40 & 1.21 & 0.17 & 0.0 & 0.33 & onemg wd + 2 & 0.02 & 12 & 11 & 4.0 & 5 & b+bb & 104.4 & 1.48 & 1.24 & 0.20 & 0.0 & 0.35 & onemg wd + 3 & 0.02 & 13 & 11 & 5.0 & 5 & b+bb & 123.3 & 1.64 & 1.43 & 0.18 & 0.0 & 0.22 & sn ic + 4 & 0.02 & 14 & 12 & 3.0 & 5 & a+ab+bb & 118.5 & 1.33 & 1.09 & 0.22 & 0.0 & @xmath155 & onemg wd + 5 & 0.02 & 14 & 12 & 5.0 & 5 & b+bb & 30.7 & 2.97 & 1.66 & 1.24 & 1.9(-4 ) & 0.25 & snib + 6 & 0.02 & 16 & 14 & 2.0 & 5 & a : contact & & & & & & & + 7 & 0.02 & 16 & 14 & 3.0 & 5 & a+ab+abb & 101.8 & 1.54 & 1.33 & 0.17 & 0.0 & 0.39 & onemg wd + 8 & 0.02 & 16 & 14 & 4.0 & 5 & b+bb & 26.2 & 3.66 & 2.05 & 1.47 & 4.5(-3 ) & 0.24 & snib + 9 & 0.02 & 16 & 14 & 5.0 & 5 & b+bb & 33.7 & 3.65 & 2.04 & 1.47 & @xmath156 & 0.25 & snib + 10 & 0.02 & 18 & 12 & 3.0 & 5 & a+ab+abb & 27.9 & 2.66 & 1.58 & 1.01 & @xmath157 & 0.26 & snib + 11 & 0.02 & 18 & 12 & 3.0 & 10 & a+ab+abb & 27.3 & 2.74 & 1.59 & 1.08 & @xmath157 & 0.26 & snib + 12 & 0.02 & 18 & 12 & 5.0 & 10 & b : contact & & & & & & & + 13 & 0.02 & 18 & 17 & 3.0 & 10 & a+ab+abb & 36.2 & 3.03 & 1.68 & 1.27 & @xmath158 & 0.25 & sn ib + 14 & 0.02 & 18 & 17 & 4.0 & 10 & a+ab & 29.7 & 3.79 & 2.14 & 1.49 & @xmath159 & 0.25 & sn ib + 15 & 0.02 & 18 & 17 & 4.0 & 10 & a+ab & 24.0 & 3.97 & 2.27 & 1.53 & @xmath160 & 0.26 & snib + 16 & 0.02 & 18 & 17 & 4.0 & 10 & a+ab & 50.0 & 3.80 & 2.16 & 1.50 & @xmath161 & 0.06 & snib + 17 & 0.02 & 18 & 17 & 4.0 & 10 & a+ab & 25.2 & 3.84 & 2.18 & 1.50 & @xmath162 & 0.26 & snib + 18 & 0.02 & 18 & 17 & 4.0 & 10 & a+ab & 30.6 & 3.73 & 2.14 & 1.43 & @xmath163 & 3.57 & snib + 19 & 0.02 & 18 & 17 & 5.0 & 3 & b & 33.1 & 3.73 & 2.33 & 1.23 & @xmath157 & 0.25 & snib + 20 & 0.02 & 18 & 17 & 5.0 & 5 & b & 32.4 & 4.04 & 2.45 & 1.4 & 0.00 & 0.33 & snib + 21 & 0.02 & 18 & 17 & 5.0 & 10 & b+bb & 31.5 & 4.41 & 2.51 & 1.68 & @xmath164 & 0.26 & snib + 22 & 0.02 & 18 & 17 & 6.0 & 10 & b & 39.3 & 4.39 & 2.56 & 1.62 & @xmath165 & 0.26 & snib + 23 & 0.02 & 25 & 19 & 6.0 & 10 & b : contact & & & & & & & + 24 & 0.02 & 25 & 24 & 2.0 & 10 & a : contact & & & & & & & + 25 & 0.02 & 25 & 24 & 3.0 & 3 & a+ab & 22.7 & 3.70 & 2.46 & 0.98 & 0.0 & 0.24 & snib + 26 & 0.02 & 25 & 24 & 3.0 & 5 & a+ab & 22.4 & 4.33 & 2.80 & 1.30 & 0.0 & 0.25 & snib + 27 & 0.02 & 25 & 24 & 3.0 & 10 & a+ab & 21.3 & 5.07 & 3.17 & 1.67 & 0.0 & 0.25 & snib + 28 & 0.02 & 25 & 24 & 3.0 & 10 & a+ab & 18.9 & 5.08 & 3.19 & 1.66 & 0.0 & 0.32 & snib + 29 & 0.02 & 25 & 24 & 4.0 & 5 & a + ab & 21.5 & 4.45 & 2.91 & 1.22 & 0.0 & 0.26 & snib + 30 & 0.02 & 25 & 24 & 6.0 & 10 & b & 27.4 & 6.49 & 4.45 & 1.63 & 0.0 & 0.39 & snib + 31 & 0.02 & 60 & 40 & 7.0 & 3 & a & 16.8 & 4.95 & 3.70 & 0.25 & 0 & 0.24 & snic + 32 & 0.004 & 16 & 12 & 3.0 & 5 & b+bb & 64.75 & 3.91 & 2.22 & 1.54 & @xmath166 & 0.24 & snib + 33 & 0.004 & 16 & 14 & 3.0 & 5 & b+bb & 19.6 & 3.90 & 2.21 & 1.53 & @xmath166 & 0.24 & snib + 34 & 0.004 & 16 & 14 & 5.0 & 5 & b+bb & 21.8 & 3.84 & 2.19 & 1.51 & @xmath167 & 0.24 & snib + 35 & 0.004 & 18 & 12 & 5.0 & 5 & b+bb & 14.4 & 4.64 & 2.76 & 1.68 & @xmath168 & 0.31 & snib + 36 & 0.004 & 18 & 12 & 8.0 & 5 & b+bb & 24.33 & 4.56 & 2.68 & 1.67 & @xmath169 & 0.33 & snib + 37 & 0.004 & 18 & 17 & 3.0 & 5 & a+ab & 18.4 & 4.42 & 2.55 & 1.67 & @xmath170 & 0.26 & snib + 38 & 0.004 & 18 & 17 & 3.0 & 10 & a+ab & 14.4 & 4.58 & 2.61 & 1.77 & @xmath171 & 0.26 & snib + 39 & 0.004 & 18 & 17 & 6.0 & 5 & b+bb & 27.0 & 4.57 & 2.71 & 1.65 & @xmath166 & 0.27 & snib + 40 & 0.004 & 25 & 12 & 3.0 & 5 & a : contact & & & & & & & + 41 & 0.004 & 25 & 12 & 6.0 & 5 & b : contact & & & & & & & + 42 & 0.004 & 25 & 19 & 3.0 & 5 & a+ab & 10.2 & 7.09 & 4.87 & 2.03 & @xmath172 & 0.32 & snib + 43 & 0.004 & 25 & 24 & 3.0 & 5 & a+ab & 13.0 & 7.31 & 5.05 & 2.07 & @xmath173 & 0.28 & snib + 44 & 0.004 & 25 & 24 & 6.0 & 5 & b : contact & & & & & & + 45 & 0.004 & 40 & 30 & 4.0 & 5 & a+ab & 9.31 & 12.0 & 9.42 & 1.24 & @xmath174 & 0.56 & bh cccccccc 5 & 2.8 & 2.73 & 1.51 & 1.16 & 163.54 & 5.08 & 0.01 + 5 & 3.0 & 2.91 & 1.60 & 1.22 & 33.40 & 1.79 & 0.03 + 5 & 4.0 & 3.72 & 2.15 & 1.43 & 8.01 & 0.73 & 0.18 + 5 & 6.0 & 4.23 & 2.77 & 1.19 & 4.19 & 0.38 & 0.25 + 5 & 8.0 & 4.95 & 3.40 & 0.85 & 3.07 & 0.30 & 0.29 + 5 & 10.0 & 5.49 & 3.93 & 0.54 & 1.59 & 0.22 & 0.47 + 5 & 15.0 & 6.44 & 4.85 & 0.31 & 0.59 & 0.18 & 0.55 + 5 & 20.0 & 7.09 & 5.44 & 0.22 & 0.48 & 0.16 & 0.66 + 10 & 3.0 & 2.95 & 1.60 & 1.27 & 34.96 & 1.84 & 0.03 + 10 & 4.0 & 3.84 & 2.20 & 1.49 & 6.94 & 0.70 & 0.15 + 10 & 6.0 & 4.93 & 3.14 & 1.55 & 4.29 & 0.51 & 0.19 + 10 & 8.0 & 6.06 & 4.13 & 1.41 & 2.55 & 0.35 & 0.33 + 10 & 10.0 & 7.01 & 4.96 & 1.12 & 1.94 & 0.25 & 0.55 + 10 & 15.0 & 8.89 & 6.80 & 0.47 & 0.54 & 0.20 & 0.82 + 10 & 20.0 & 10.33 & 8.21 & 0.31 & 0.40 & 0.16 & 1.17 | we investigate the evolution of type ib / c supernova ( sn ib / c ) progenitors in close binary systems , using new evolutionary models that include the effects of rotation , with initial masses of 12 25 @xmath0 for the primary components , and of single helium stars with initial masses of 2.8 20 @xmath0 .
we find that , despite the impact of tidal interaction on the rotation of primary stars , the amount of angular momentum retained in the core at the presupernova stage in different binary model sequences converge to a value similar to those found in previous single star models .
this amount is large enough to produce millisecond pulsars , but too small to produce magnetars or long gamma - ray bursts .
we employ the most up - to - date estimate for the wolf - rayet mass loss rate , and its implications for sn ib / c progenitors are discussed in detail . in terms of stellar structure , sn ib / c progenitors in binary systems at solar metallicity are predicted to have a wide range of final masses up to about 7 @xmath0 , with helium envelopes of @xmath1 .
our results indicate that , if the lack of helium lines in the spectra of sne ic were due to small amounts of helium ( e.g. @xmath2 ) , the distribution of both initial and final masses of sn ic progenitors should be bimodal .
furthermore , we find that a thin hydrogen layer ( @xmath3 ) is expected to be present in many sn ib progenitors at the presupernova stage .
we show that the presence of hydrogen , together with a rather thick helium envelope , can lead to a significant expansion of some sn ib / c progenitors by the time of supernova explosion .
this may have important consequences for the shock break - out and supernova light curve .
we also argue that some sn progenitors with thin hydrogen layers produced via case ab / b transfer might be related to type iib supernova progenitors with relatively small radii of about @xmath4 . |
it is considered that the results coming from atmospheric neutrino experiments @xcite , as well as solar neutrino experiments @xcite , are strong evidence of neutrino oscillation indicating the presence neutrino mass and flavor mixing . however , several alternative explanations to these results , which do not invoke neutrino mass and/or mixing , exist and are not yet excluded @xcite . the interesting idea that gravitational forces may induce neutrino mixing and flavour oscillations , if the weak equivalence principle of general relativity is violated ,
was proposed about a decade ago @xcite , and thereafter , many works have been performed on this subject @xcite .
many authors have investigated the possibility of solving the solar neutrino problem ( snp ) by such gravitationally induced neutrino oscillations @xcite , generally finding it necessary , in this context , to invoke the msw like resonance @xcite in order reduce appropriately the solar neutrino fluxes with the specific energy dependence , required to explain the data .
nevertheless we have demonstrated in ref .
@xcite that all the recent solar neutrino data coming from gallium , chlorine and water cherenkov detectors can be well accounted for also by long - wavelength neutrino oscillations induced by a violation of the equivalence principle ( vep ) . in this talk , we present updated fit of such gravitationally induced long - wavelength oscillation solution @xcite to the most recent solar neutrino data which include the first gno measurement @xcite .
we follow the framework proposed in refs . @xcite . in the presence of violation of equivalence principle ,
neutrino mixing and oscillation can occur even if neutrinos are massless . in this work
we assume oscillations only between two species of neutrinos , which are massless ( or degenerate in mass ) , either between active and active ( @xmath0 ) or active and sterile ( @xmath1 , @xmath2 being an electroweak singlet ) neutrinos .
to describe the vep induced massless neutrino oscillation mechanism , phenomenologically , we can simply do the following replacement in the usual mass induced oscillation formula : @xmath3 and @xmath4 , where @xmath5 is the mass squared difference , @xmath6 is the gravitational potential which is assume to be constant in our work , @xmath7 is the usual mixing angle which relate weak and mass eigenstates and @xmath8 is the mixing which relates weak and gravitational eigenstates , and @xmath9 is the quantity which measures the magnitude of vep .
the survival probability of @xmath10 produced in the sun after traveling the distance @xmath11 to the earth is given by , @xmath12 where the oscillation wavelength @xmath13 for a neutrino with energy @xmath14 is given by @xmath15 \left[\frac{10^{-15 } } { |\phi \delta \gamma|}\right ] \left[\frac { { \mbox{mev}}}{e}\right ] , \label{wavelength}\ ] ] which in contrast to the wavelength for mass induced neutrino oscillations in vacuum , is inversely proportional to the neutrino energy .
this energy dependence is very crucial in obtaining a good fit to the total rates without causing any problem with the sk spectrum @xcite .
we present here the results of our analysis only for active to active conversion since the results are qualitatively similar for active to sterile one @xcite .
-1.4 cm -0.7 cm -2.3 cm fig.1 : parameter region of @xmath16 and @xmath17 allowed by the total rates only for @xmath18 .
the best fit point is indicated by the filled circle .
[ fig1 ] 0.3 cm we performed the same statistical analysis as we did for three flavor vaccum oscillation solution to the snp in ref .
@xcite , which is slightly different from our analysis in ref .
we fit vep parameters @xmath17 and @xmath8 , and an extra normalization factor @xmath19 for the @xmath20b neutrino flux , to the most recent experimental results coming from homestake @xcite , sage @xcite , gallex @xcite and gno @xcite combined , and super - kamiokande ( sk ) @xcite .
we present in fig . 1 the allowed region determined only by the total rates for fixed @xmath20b flux .
we then present in fig .
2 the result for spectral shape analysis fitting only the @xmath20b spectrum measured by sk @xcite in the same way as we did in ref . @xcite .
finally , we perform a combined fit of the rates and the spectrum obtaining the allowed region presented in fig . 3 .
the combined allowed region is essentially the same as the one obtained by the rates alone .
we also show in fig .
4 , the predicted spectra for the best fitted parameters , which are in good agreement with the data .
-1.5 cm -0.5 cm -2.4 cm fig .
2 : same as in fig . 1 but for sk recoil electron spectrum .
it is the inner part of the contours which is excluded by the data .
-1.5 cm [ fig2 ] -0.5 cm -2.4 cm fig .
3 : same as in fig . 1 but for the rates and sk spectrum combined .
-0.3 cm [ fig3 ] -1.6 cm -0.3 cm -2.5 cm fig .
4 : expected spectrum for the best fitted parameters .
0.2 cm best fitted parameters as well as @xmath21 can be found in table i. from this table , we can conclude that the fit is quite good . for sterile conversion
, we found somewhat worse , but acceptable , fit compared to the active one .
for example , we found for the combined fit , @xmath21 = 19.5 for 24 d.o.f . for fixed @xmath19 .
-0.2 cm 0.2 cm . the best fitted vep parameters and values of @xmath21 .
values of @xmath17 are shown in units of @xmath22 .
number of degree of freedom are from top to bottom , 2 , 1 , 15 , 24 and 23 .
@xmath19 is fixed unless it is indicated as `` free '' in the parenthese .
we include 6 bins for the sk zenith angle dependence in the combined fit .
[ cols="^,^,^,^,^",options="header " , ] [ tab1 ] -1.2 cm
we have obtained a very good fit to the most recent solar neutrino data for the vep induced long - wavelength neutrino oscillation .
let us finally remark that , in contrast to the usual vacuum oscillation solution to the snp , in this vep scenario no strong seasonal effect is expected in any of the present or future experiments , even the ones that will be sensitive to @xmath23be neutrinos .
see ref .
@xcite for more detail .
hn thanks e. lisi for a useful discussion during the now2000 workshop .
this work was supported by the brazilian funding agencies fapesp and cnpq .
v. gavrin for the sage collab . , talk given at the xix international conference on neutrino physics and astrophysics ( neutrino 2000 ) , sudbury , canada , june 16 - 21 , 2000 , available at http://nu2000.sno.laurentian.ca/. w. hampel _ et al .
_ , phys .
b * 447 * ( 1999 ) 127 .
j. pantaleone , a. halprin , and c. n. leung , phys .
d * 47 * ( 1993 ) r4199 ; j. n. bahcall , p. i. krastev , and c. n. leung , phys .
d * 52 * ( 1995 ) 1770 ; s. w. mansour and t. k. kuo , phys .
d * 60 * ( 1999 ) 097301 . | we have updated the analysis for the solution to the solar neutrino problem by the long - wavelength neutrino oscillations induced by a tiny breakdown of the weak equivalence principle of general relativity , and obtained a very good fit to all the solar neutrino data . |
among all nonperturbative approaches , the qcd sum rules method @xcite , is an especially powerful one for calculating the fundamental parameters of hadrons . in this approach a deep connection
is established between the hadronic parameters and the qcd vacuum via a few condensates .
this method has been applied very successfully to the various problems of hadron physics and is discussed in detail in many review articles @xcite@xcite .
one of the important static characteristic parameters of hadrons is their magnetic moment .
nucleon magnetic moment was investigated in the framework of the traditional qcd sum rules method in @xcite . within this method , the @xmath7 transition magnetic moment was calculated in @xcite .
later this method was used in determining magnetic moments of baryons containing heavy mesons @xcite . in the present
work our goal is to calculate @xmath4 ( @xmath2 or @xmath3 ) transition magnetic moment in the framework of the light cone qcd sum rules method ( lcqsr ) ( more about this approach can be found in @xcite and @xcite and the references therein ) , which is an alternative approach to the traditional qcd sum rules method .
note that magnetic moments of the nucleons and decuplet baryons were studied in @xcite using the lcqsr approach .
moreover , magnetic moments of heavy @xmath1 baryons and @xmath8 transition magnetic moment were investigated in @xcite and @xcite , respectively , using the lcqsr method .
the paper is organized as follows : in section 2 , the lcqsr for @xmath4 transition magnetic moment is derived .
we present numerical calculations and conclusion in section 3 .
in order to calculate @xmath4 transition magnetic moment we start by considering the following correlator function : [ pi1 ] = i d^4 x e^i p x 0 t\{__q(x ) |__q(0 ) } 0 _ f _ , where @xmath9 is the external electromagnetic field , @xmath10 and @xmath11 are the interpolating currents with @xmath0 and @xmath1 quantum numbers , respectively .
it is well known that there is a continuum of choices for the baryon interpolating currents .
the general form of @xmath0 and @xmath1 currents can be written as @xcite [ eta ] _ _ q 2 _ _ 1 + b _ _ 2 , + _ _ q 2 _
_ 1 + b^__2 , where @xmath12 and @xmath13 are arbitrary parameters and [ eta1 ] _ _ 1 _ abc , + [ eta2 ] _ _ 2 _ abc , + [ eta3 ] _ _ 1 _ abc , + [ eta4 ] _ _ 2 _ abc , with @xmath14 , @xmath12 , and @xmath3 being the color indices .
note that the ioffe current corresponds to the choice @xmath15 .
phenomenological part of the sum rules is obtained by inserting a complete set of states with @xmath0 and @xmath1 quantum numbers between the currents in eq .
( [ pi1 ] ) , which can be expressed as [ pi2 ] _
q(p_1 ) _ q ( p_2 ) _ f _ + _ h_i , h_i h_i(p_1 ) h_i ( p_2 ) _ f _ , where @xmath16 , and @xmath17 is the photon momentum .
the second term in eq . ( [ pi2 ] ) takes higher resonances and continuum contributions into account .
the coupling of currents with the corresponding baryon states is parametrized by the overlap amplitudes @xmath18 which are defined as [ amp1 ] 0 _ _ q _ q _ _ q u__q ( p ) , + 0 _ _ q _ q _ _ q u__q ( p ) .
it follows from eq .
( [ pi2 ] ) that in calculating the phenomenological part of the correlator , we need the expression of the matrix element @xmath19 , which can be parametrized as [ amp2 ] _
q(p_2 ) _ f _ |u ( p_1 ) u ( p_2 ) ^ , + @xmath20 are the form factors which are functions of @xmath21 , and @xmath22 is the polarization vector of photon . in order to calculate @xmath23 transition magnetic moment only the values of the form factors at @xmath24 are needed , since the photon is a real one . using eqs .
( [ pi2])([amp2 ] ) , for the hadronic representation of the correlator function , we get [ pi3 ] = - _ _ q _ _ q ^ . among a number of different structures which appear in the phenomenological part , we choose the structure @xmath25 that contains transition magnetic form factor @xmath26 .
this form factor gives transition magnetic moment in units of @xmath27 at @xmath24 .
choosing the above mentioned structure from the phenomenological part , we get [ pi4 ] = - _ _ q _ _ q _ _ q _ q , where @xmath28 is the transition magnetic moment . in order to construct sum rules , theoretical part of the correlator function needs to be calculated .
calculation of the correlator function in qcd leads to the following result : [ cor ] + & & 2 5 5 + 2 b 5 5 + 2 b 5 5 + 2 b b 5 5 + 5 5 + b 5 5 + b 5 5 + b b 5 5 - 5 5 + b 5 5 -b 5 5 -b b 5 5 + 2 5 5 -2b 5 5 -2 b 5 5 + 2 b b 5 5 + 5 5 + b 5 5 + b 5 5 + b b 5 5 - 5 5 + b 5 5 -b 5 5 -b b 5 5 0 , where @xmath29 with @xmath30 being the charge conjugation operator , @xmath31 denotes transpose of the operator and @xmath32 is the quark propagator with the subindices referring to the corresponding quarks .
theoretical part of the correlator contains two pieces : perturbative and nonperturbative .
perturbative part corresponds to the case when photon is radiated from the freely propagating quarks .
its expression can be obtained by making the following replacement in each one of the quark propagators and keeping the other two as they are in eq .
( [ cor ] ) [ sss ] s^ab _ - ( dy f^ y_s^free ( x - y ) _
s^free(y ) ) ^ab _ , where the fock
schwinger gauge @xmath33 has been used and @xmath34 is the free quark operator . in @xmath35representation
the propagator of the free massive quark is [ fmq ] s_q^free = - i k_2 m_q , where @xmath36 is the heavy quark mass and @xmath37 are bessel functions . the expression for massless propagator @xmath38 can be obtained from eq .
( [ fmq ] ) by making use of the following expansions for bessel functions k_1(x ) & ~ & + o ( x ) , + k_2(x ) & ~ & - + o ( x^2 ) , and then formally setting @xmath39 .
the nonperturbative contributions can be obtained from eq .
( [ cor ] ) by making the following replacement in each one of the massless quark propagators : [ rep ] s^ab _ - |q^a a_j q^b a_j _ , where @xmath40 and sum over @xmath41 is implied , and the other two propagators are the full propagators involving both perturbative and nonperturbative contributions . in order to calculate perturbative and nonperturbative contributions ,
the explicit expressions of the heavy and light quark propagators in the presence of external field are needed . the complete light cone expansion of the propagator in external field is presented in @xcite .
it receives contributions from the nonlocal operators @xmath42 , @xmath43 , @xmath44 , where @xmath45 is the gluon field strength tensor .
here we consider operators with only one gluon field and neglect terms with two gluons @xmath46 , and four quarks @xmath44 .
neglecting these terms can be justified on the basis of an expansion in conformal spin @xcite . in this approximation
massive and massless quark propagators are given by the following expressions , [ hvy ] i s_q ( x ) i s_q^free ( x ) -
i g_s e^-ikx _ 0 ^ 1 dv , + + + [ lig ] i s_q ( x ) i s_q^free(x ) - 1 + + i g_s dv , where @xmath47 and @xmath48 are the free propagators of the heavy and light quarks , respectively , and @xmath49 is defined from the relation |q
^ q = m_0 ^ 2 . in eq .
( [ lig ] ) we neglect the operators with dimensions larger than five , since their contributions are negligible . as can be seen from eqs .
( [ cor])([lig ] ) , in order to calculate the theoretical part of the correlator function ( [ pi1 ] ) , the matrix elements @xmath50 of the non local operators between photon and vacuum states are needed .
up to twist4 these matrix elements are defined in terms of the photon wave functions in the following way @xcite@xcite : [ pwf ] ( q ) |q _ _ 5 q 0 e_q _ ^q^x^_0 ^ 1 du e^i u qx ( u ) , + ( q ) |q _ q 0 i e_q _ 0 ^ 1 du e^i u q x \ { ( _ q_- _ q _ ) + g_2 ( u ) } . in eq . ( [ pwf ] ) , @xmath51 is the magnetic susceptibility of the quark condensate , @xmath52 is the quark charge , @xmath53 and @xmath54 are the leading twist2 photon wave functions , while @xmath55 and @xmath56 are the twist4 functions . theoretical part of the correlator ( [ pi1 ] ) can be obtained by substituting photon wave functions and expressions for massive and massless quark propagators into eq .
( [ cor ] ) .
sum rules for @xmath23 transition magnetic moment are obtained by equating the phenomenological and theoretical parts of the correlator ( [ pi1 ] ) . performing double borel transformations on the variables @xmath57 and @xmath58 on both sides of the correlator , which suppress the continuum and higher state contributions ( see @xcite for details ) , @xmath23 transition magnetic moment
can be expressed as [ mgm ] + & & ( 2 + b + b + 2 b b ) m^6 ( 2,-1,m_q^2/m^2 ) + ( -3 b + b + 2 b b ) ( u_0 ) m^4 ( 2,0,m_q^2/m^2 ) + ( 2 + b + b + 2 b b ) f ( u_0 ) m^4 ( 1,-1,m_q^2/m^2 ) + ^2 ( -1 + b)(1 + b ) + \ { 3 f_1(m_q^2/m^2 ) - } + m_0 ^ 2 f ( u_0 ) + \ { ( -2 -b -b + 4 b b ) + 3 ( -1 + b b ) f_5(m_q^2/m^2 ) } + ( -2 -b -b + 4 b b ) m^2 ( 1,0,m_q^2/m^2 ) + ( b -3 b + 2 b b ) ( g_1(u_0 ) - g_2(u_0 ) ) m^2 ( 1,0,m_q^2/m^2 ) + ( -1+b)(1+b ) ( u_0 ) m^2 f_5(m_q^2/m^2 ) + ^2 ( -1 + b)(1+b ) ( g_1(u_0 ) - g_2(u_0 ) ) f_4(m_q^2/m^2 ) + ^2 ( -1+b)(1+b ) ( u_0 ) + ( 2 + b + b -4 b b ) f ( u_0 ) f_5(m_q^2/m^2 ) + m_0 ^ 2 } , where the functions @xmath59 and @xmath60 are defined as [ psi ] ( , , z ) _
1^dt e^-t z t^--1 ( t-1)^-1 , ( > 0 ) , + f_1 ( z ) z z - 2 e^-z , + f_2 ( z ) z^2 - 4 z + 2 e^-z , + f_3 ( z ) z^2 - 6 z + 6 e^-z , + f_4 ( z ) z e^-z , + f_5 ( z ) e^-z , and u_0 = , m^2= , with @xmath61 and @xmath62 being the borel parameters .
since masses of the initial and final baryons are very close to each other , we will set @xmath63 and @xmath64 , from which it follows that @xmath65 .
it should be noted that borel transformations of bessel functions are given in @xcite .
it follows from eq .
( [ mgm ] ) that determination of @xmath66 transition magnetic moment requires a knowledge of the residues @xmath67 and @xmath68 .
these residues can be determined from heavy baryons mass sum rules ( for the @xmath69 case the residue is calculated in @xcite ) [ ml1 ] + ( 1 + 4 b - 5 b^2 ) m^4 ( 1,-1,m_q^2/m^2 ) + m_0 ^ 2 ^2 \{3 ( 5 + 2 b + 5 b^2 ) + ( -1+b)^2 f_5(m_q^2/m^2 ) } + m_0 ^ 2 m^2 + ^2 ( 5 + 2 b + 5 b^2 ) f_5(m_q^2/m^2 ) , + + [ ml2 ] + ^2 + ( 1 + 4 b - 5 b^2 ) m^2 ( 2,0,m_q^2/m^2 ) + ( -13 + 2 b + 11 b^2 ) f_5(m_q^2/m^2 ) + m_0 ^ 2 , + + [ ms1 ] + ( 1-b^2 ) m^4 ( 1,-1,m_q^2/m^2 ) + m_0 ^ 2 ^2 \ { ( 5 + 2 b + 5 b^2 ) + ( 3 + 2 b + 3 b^2 ) f_5(m_q^2/m^2 ) } + ( -1+b^2 ) m^2 + ^2 ( 5 + 2 b + 5 b^2 ) f_5(m_q^2/m^2 ) , + + [ ms2 ] + ^2 ( 1-b)^2 + ( -1 + b^2 ) m^2 ( 2,0,m_q^2/m^2 ) + ( 1-b)^2 f_5(m_q^2/m^2 ) + m_0 ^ 2 ( -1 + b^2 ) .
( [ ml1 ] ) and ( [ ms1 ] ) correspond to the structure proportional to the unit operator , while eqs .
( [ ml2 ] ) and ( [ ms2 ] ) correspond to the structure @xmath70 .
finally , we remark that subtraction of the continuum contribution in eqs .
( [ mgm])([ms2 ] ) can be performed with the help of the following replacement [ sub ] m^2n ( , , m_q^2/m^2 ) _ m_q^2^s_0 ds e^-s / m^2 _ 1^s / m_q^2 dt ( s - t m_q^2)^n-1 t^- -1 ( t-1)^- 1 , for @xmath71 and @xmath72 .
in this section we present our numerical calculation for @xmath66 transition magnetic moment .
it follows from eq .
( [ mgm ] ) that the main input parameters of the lcqsr are the photon wave functions .
the photon wave functions which we use in the present work are given as @xcite : ( u ) 6 u ( 1-u ) , ( u ) = 1 , + g_1(u ) - ( 1-u)(3-u ) , g_2(u ) = - ( 1-u)^2 .the values of the other input parameters entering to the sum rules are : @xmath73 @xcite ( in @xcite this quantity is estimated to be @xmath74 ) , @xmath75 , @xmath76 @xcite , @xmath77 , @xmath78 , and @xmath79 .
few word about the value of @xmath80 are in order .
the analysis of the magnetic susceptibility @xmath51 by the qcd sum rules ( see @xcite and @xcite ) shows that at an accuracy of the order of @xmath81 , the dominant contribution to this quantity comes from @xmath82 meson at @xmath83 .
it has been suggested that on these argument that the vector dominance model works sufficiently well for electromagnetic properties of hadrons in the constant external field limit @xmath84 . to the quoted accuracy , in this limit for the normalization constant
is obtained @xcite @xmath85 at @xmath86 and @xmath87 . the parameters @xmath12 and @xmath13 are completely arbitrary and the physical quantities should not depend on the precise value of these parameters .
to simplify further analysis , we will set @xmath88 . since transition magnetic moment is a physical quantity it must be independent of the auxiliary parameters @xmath12 , borel mass square @xmath89 and continuum threshold @xmath90 .
so , our problem reduces to determining the respective regions for which @xmath66 transition magnetic moment is independent of the above mentioned parameters .
for this aim we prefer to consider the following three steps . in the first step
, we try to find the working region of @xmath89 where @xmath66 is independent of the borel parameter @xmath89 , at fixed values of @xmath12 and @xmath90 . along these lines
, we present in figs .
( 1 ) and ( 2 ) the dependence of @xmath91 and @xmath92 on @xmath89 , respectively . from both figures
we observe that , except @xmath93 case , transition magnetic moments seem to be almost independent for the different choices of @xmath12 and @xmath90 .
the working region for @xmath91 transition magnetic moment is @xmath94 , while it is @xmath95 for the @xmath92 case .
the next step is to determine the working region for the parameter @xmath12 . for this purpose
we use the fact that both mass sum rules for @xmath1 and @xmath0 should be positive .
note also that the mass of the @xmath1 baryon can be obtained by dividing eq .
( [ ml1 ] ) with eq .
( [ ml2 ] ) , and that of @xmath0 can be obtained by dividing eq .
( [ ms1 ] ) with eq .
( [ ms2 ] ) . in order to see whether this requirement is fulfilled or not , in figs .
( 3 ) and ( 4 ) we present the dependence of the sum rules for the masses of the above mentioned baryons on @xmath96 , where @xmath97 is determined from the relation @xmath98 .
it follows from these figures that the working region of @xmath12 , which guarantees the positiveness of the sum rules , are in the intervals @xmath99 for @xmath69 baryon and @xmath100 for @xmath101 baryon .
similar analysis for the @xmath0 baryons leads to the following result for the working region of the parameter @xmath12 : @xmath102 and @xmath103 for the @xmath104 and @xmath105 baryons , respectively .
so , the common working region for the parameter @xmath12 for both cases is @xmath106 .
having decided about the restriction of the parameter @xmath12 , our third and final step is to determine the main objective of the present work , i.e. , @xmath66 transition magnetic moment .
as before , we are supposed to find a region for the parameter @xmath12 where transition magnetic moment @xmath66 is independent of its variation . in the first step of our analysis we have decided on the working region of the borel parameter @xmath89 for which @xmath66 does not depend on its variation , and additionally , we have verified its insensitiveness to several different choices of the continuum threshold @xmath90 .
( 5 ) presents the dependence of transition magnetic moment @xmath91 ( in units of the nucleon magneton @xmath107 ) on @xmath96 , at @xmath108 and at three fixed values of @xmath90 .
similarly , in fig .
( 6 ) we present the dependence of @xmath92 on @xmath96 , at @xmath109 and at three fixed values of @xmath90 . from both figures
we deduce that @xmath66 is quite stable in the region @xmath110 , and can be said to be practically independent of the parameter @xmath12 and the continuum threshold @xmath90 .
it follows from all these arguments that _ _ c_c - ( 1.5 0.4 ) _
n , + _ _ b_b - ( 1.6 0.4 ) _
n , where the uncertainty in the results can be mainly attributed to the variation of the continuum threshold @xmath90 , borel parameter @xmath89 and the twist3 photon wave functions which are neglected , since we estimate that their contribution to the transition magnetic moment is less than 5% , as well as to the uncertainties in the values of the parameters @xmath80 , @xmath51 , and @xmath111 . finally , we compare our results on @xmath112 transition magnetic moment with the ones predicted by other methods in literature .
transition magnetic moment @xmath92 is calculated in the lcqsr to leading order in heavy quark effective theory which predicts @xmath113 @xcite .
this result is calculated using ioffe currents for @xmath105 and @xmath101 .
when we compare our results with the results given in @xcite we see that the discrepancy between them is about 50% . in our opinion
this discrepancy can be attributed to the fact that the result presented in @xcite is calculated for the choice @xmath93 ( ioffe current ) , which is unphysical in our case . in summary , transition magnetic moments
@xmath112 are calculated in the framework of the lcqsr , using the general form of the interpolating currents for @xmath0 and @xmath1 mesons . in our analysis , only two particle photon wave functions are taken into account while three particle photon wave functions are neglected , since we estimate that their contribution to the transition magnetic moments , as has already been mentioned above , is less than 5% . 99 m. a. shifman , v. i. vainshtein , v. i. zakharov , _ nucl .
phys . _ * b147 * ( 1979 ) 385 .
l. j. reinders , h. rubinstein and s. yazaki , _ phys .
* 127 * ( 1985 ) 1 .
vacuum structure and qcd sum rules , ed : m. a. shifman , north holland , amsterdam ( 1992 ) .
b. l. ioffe , in `` the spin structure of nucleon '' ed . by b. frois ,
w. w. hughes , n. de croot , world scientific ( 1997 ) , prep : hep ph/9511401 .
m. a. shifman , _ prog .
* 131 * ( 1998 ) 1 .
s. narison , `` qcd spectral sum rules '' , world scientific ( 1989 ) . p. colangelo and a. khodjamirian , in `` at the frontier of particle physics / handbook of qcd '' , ( ed . ) m. shifman , volume 3 , page 1495 ( 2001 ) .
b. l. ioffe , and a. v. smilga , _ nucl .
phys . _ * b232 * ( 1984 ) 109 .
i. i. balitsky and a. v. yung , _ phys .
lett . _ * b129 * ( 1983 ) 328 .
b. l. ioffe and m. smilga , _ phys .
_ * b133 * ( 1983 ) 436 ; + s
zhu , w. y. p. hwang and z. s. yang , _ phys .
rev . _ * d51 * ( 1998 ) 1527 .
zhu , w. y. p. hwang and z. s. yang , _ phys .
* d56 * ( 1997 ) 7273 .
v. m. braun , prep : hep ph/9801222 ( 1998 ) .
v. m. braun , i. e. filyanov , _
_ * c44 * ( 1989 ) 157 .
t. m. aliev , a. zpineci , m. savci , _ nucl .
phys . _ * a678 * ( 2000 ) 443 ; + t. m. aliev , a. zpineci , m. savci , _ phys . rev .
_ * d62 * ( 2000 ) 053012 .
t. m. aliev , a. zpineci , m. savci , _ phys .
* d65 * ( 2002 ) 056008 .
t. m. aliev , a. zpineci , m. savci , _ phys .
lett . _ * b516 * ( 2001 ) 299 .
h. g. dosch , m. jamin and s. narison , _ phys .
lett . _ * b220 * ( 1989 ) 251 . v. chung , h. g. dosch , m. kremer , d. scholl , _ nucl .
phys . _ * b197 * ( 1982 ) 55 .
i. i. balitsky , v. m. braun , _ nucl .
_ * b311 * ( 1988 ) 541 .
v. m. braun , i. e. filyanov , _
z. phys . _ * c48 * ( 1990 ) 239 .
i. i. balitsky , v. m. braun , a. v. kolesnichenko , _ nucl .
* b312 * ( 1989 ) 509 .
a. ali , v. m. braun , _ phys .
* b359 * ( 1995 ) 223 .
v. m. belyaev , v. m. braun , a. khodjamirian and r. rckl , + _ phys .
* d51 * ( 1995 ) 6177 .
t. m. aliev , m. savci , _ phys .
rev . _ * d61 * ( 2000 ) 0160008 . v. m. belyaev , b. yu .
_ z. phys . _
* c30 * ( 1986 ) 151 .
v. m. belyaev , ya .
i. kogan , _ yad .
_ * 40 * ( 1984 ) 1035 .
i. i. balitsky , a. v. kolesnichenko , _ yad .
* 41 * ( 1985 ) 282 .
v. m. belyaev , b. l. ioffe , _ jetp _ * 56 * ( 1982 ) 493 .
shi lin zhu and yuan ben dai , _ phys .
_ * d59 * ( 1999 ) 114015 .
* fig . ( 1 ) * the dependence of the transition magnetic moment @xmath91 on @xmath89 at two different values of the continuum threshold @xmath114 and @xmath115 , for several fixed values of the parameter @xmath12 . here in this figure and figs .
( 2 ) , ( 5 ) and ( 6 ) the transition magnetic moments are given in units of the nucleon magneton @xmath107 .
+ + * fig .
( 2 ) * the dependence of the transition magnetic moment @xmath92 on @xmath89 at two different values of the continuum threshold @xmath116 and @xmath117 , for several fixed values of the parameter @xmath12 .
+ + * fig .
( 3 ) * the dependence of the mass sum rule @xmath118 on @xmath119 , at @xmath108 and at three different values of the continuum threshold @xmath120 and @xmath121 .
+ + * fig .
( 4 ) * the dependence of the mass sum rule @xmath122 on @xmath119 , at @xmath109 and at three different values of the continuum threshold @xmath123 and @xmath124 .
+ + * fig .
( 5 ) * the dependence of the transition magnetic moment @xmath91 on @xmath119 , at the fixed value @xmath108 of the borel parameter , and at three different values of the continuum threshold @xmath120 and @xmath121 .
+ + * fig .
( 6 ) * the dependence of the transition magnetic moment @xmath92 on @xmath119 , at the fixed value @xmath109 of the borel parameter , and at three different values of the continuum threshold @xmath123 and @xmath124 . | using the general form of @xmath0 and @xmath1 ( @xmath2 or @xmath3 ) currents , @xmath4 transition magnetic moments are calculated in framework of the light cone qcd sum rules method . in this approach
nonperturbative effects are described by the photon wave functions and only two particle photon wave functions are taken into account .
our predictions on transition magnetic moments are @xmath5 and @xmath6 .
a comparison of our results with the ones existing in the literature is given .
16.3 true cm 23.0 true cm -0.8 true in 0.00 true in 0^*0 5_5 s_s^ ? ?
o _ ^0
_ pacs number(s ) : 11.55.hx , 13.40.em , 14.20.lq , 14.20.mr |
given a positive integer @xmath1 , a global field @xmath2 containing all @xmath3th roots of unity , and an irreducible , reduced root system @xmath4 of rank @xmath5 , weyl group multiple dirichlet series are dirichlet series in @xmath5 complex variables , with analytic continuation to @xmath6 and a group of functional equations isomorphic to the weyl group of @xmath4 .
such series have applications in analytic number theory and arise , for example , in the study of moments of @xmath7-functions @xcite .
we will be concerned with a class of weyl group multiple dirichlet series that arise as fourier - whittaker coefficients of metaplectic eisenstein series ( eisenstein series on covers of reductive groups ) . because the coefficients involve gauss sums , in general , such series are _ not _ eulerian .
however , they do satisfy a twisted analogue of an euler product : each prime @xmath0 of @xmath2 corresponds to a local factor , and like classical euler factors , the local @xmath0-parts are generating functions for the @xmath0-power coefficients of the series .
in particular , the @xmath0-parts completely determine the global series . due to the complexity inherent in computing on metaplectic groups
, significant effort has been devoted to the combinatorial problem of how to define @xmath0-parts in a way such that the resulting series coincide with those coming from eisenstein series . at present
, there are two main approaches for which the resulting series have been shown to yield the desired analytic properties of meromorphic continuation and weyl group of functional equations .
the first of brubaker , bump , and friedberg defines the @xmath0-parts term by term via gelfand - tsetlin patterns , crystal graphs , or related combinatorial devices .
this approach produces the desired analytic properties for all @xmath4 when @xmath1 is sufficiently large @xcite , for all @xmath1 when @xmath8 @xcite , and in various other cases for classical types , e.g. @xcite .
the second of chinta and gunnells defines the @xmath0-parts all at once with an averaging technique analogous to the weyl character formula .
this approach produces a global object with the desired analytic properties for all @xmath4 and all @xmath1 @xcite . in this paper
we study the structure of @xmath0-parts of weyl group multiple dirichlet series defined using the chinta gunnells technique @xcite .
our main goal is to show that the methods of @xcite and @xcite give rise to the same series , in the situation where both are applicable . for @xmath8 and @xmath9
the result follows from tokuyama s formula @xcite , which expresses a deformation of a weyl character as a sum over a crystal graph . for @xmath8 and general @xmath1 , the result follows from the the combined works of chinta and offen @xcite and mcnamara @xcite , who compare the two methods via local whittaker functions . for @xmath4 simply laced and @xmath10 , chinta , friedberg , and gunnells @xcite provide further evidence by showing that the stable ( see section [ sec : stable ] ) coefficients of the local @xmath0-parts agree .
we extend the results of @xcite to all @xmath4 and @xmath1 . our main tool to understand the structure of the @xmath0-parts is an explicit set of recurrence relations on the coefficients , which we state in theorem [ thm : rec ] .
these relations allow us to prove theorem [ thm : support ] , which extends ( * ? ? ?
* theorem 3.2 ) on the support of the @xmath0-parts to all @xmath4 and all @xmath1 . with theorem [ thm : support ]
, we compare the stable coefficients of @xcite and @xcite ; theorem [ thm : stable ] extends ( * ? ? ?
* theorem 4.1 ) and shows they do indeed agree .
finally , theorem [ thm : unstab ] addresses the extent to which the recurrence relations uniquely determine the @xmath0-parts .
in addition to analyzing the @xmath0-parts , we explore the structure of modified @xmath0-parts ( obtained by multiplying the @xmath0-parts with a rational factor ) that have applications to the case when @xmath2 is a function field over a finite field .
when @xmath2 is the rational function field , up to a variable change the modified @xmath0-parts coincide with the global weyl group multiple dirichlet series .
this phenomenon was first noticed in @xcite and mirrors the similarity between the zeta function of the rational function field and its euler factors . as a complement to theorem [ thm : support ] , theorem
[ thm : gap ] describes the support of these modified @xmath0-parts . just as in the case of the zeta function of a function field
, we expect the associated multiple dirichlet series to encode much information about the arithmetic of the defining curve .
the first step is to understand the support of the series , which author will pursue in a forthcoming paper .
see also @xcite .
this paper proceeds as follows . in section
[ sec : prelims ] , we review the chinta gunnells construction of the @xmath0-parts @xcite . in section [ sec : rec ] , we derive an explicit set of recurrence relations on the coefficients of the @xmath0-parts , allowing us to prove theorem [ thm : support ] on the support of the @xmath0-parts . at the end of section [ sec : rec ]
we also prove theorem [ thm : gap ] on the support of the modified @xmath0-parts .
section [ sec : stable ] compares the stable coefficients of @xcite and @xcite , and section [ sec : unstable ] details what is known about the unstable coefficients and the extent to which the @xmath0-parts are uniquely determined .
i thank paul e. gunnells , who advised the thesis that led to this work for invaluable advice and guidance during this project .
i also thank icerm and the organizers and participants of the spring 2013 semester on automorphic forms , combinatorial representation theory and multiple dirichlet series for their support during the final stages of my dissertation . in particular , thank you to gautam chinta , ben brubaker , solomon friedberg , daniel bump , jeff hoffstein , anna pusks , ian whitehead , and tingfang lee for helpful conversations .
the @xmath0-parts of weyl group multiple dirichlet series are built from combinatorial and number theoretic data . in this section ,
we review essential definitions related to root systems and gauss sums and then describe the construction of the @xmath0-parts .
let @xmath4 be an irreducible , reduced root system of rank @xmath5 and @xmath11 the simple roots .
the root lattice @xmath12 of @xmath4 is the @xmath13-span of the simple roots . for @xmath14 , define the generalized height function @xmath15 as @xmath16 denote by @xmath17 the weyl group of @xmath4 , and choose a @xmath17-invariant symmetric bilinear inner product @xmath18 on @xmath19 normalized so that all short roots have length one , in which the short roots are taken to have length 1 . ] .
this normalization implies that for any @xmath20 , we have @xmath21 . for @xmath20 , further define @xmath22 . then the simple reflections @xmath23 generate @xmath17 and the cartan matrix @xmath24 of @xmath4 is defined by @xmath25 .
define the simple coroots @xmath26 for @xmath27 .
the dual lattice to @xmath12 with respect to the coroot basis is the weight lattice @xmath7 of @xmath4 whose basis is given by the fundamental weights @xmath28 defined by @xmath29 .
there is a partial order on the weights : for @xmath30 , we say @xmath31 if @xmath32 with all @xmath33 nonnegative .
the weight @xmath34 is said to be dominant if @xmath35 for all @xmath27 and strongly dominant if the inequality is strict .
for instance @xmath36 is strongly dominant .
let @xmath1 be a positive integer and @xmath2 a global field containing all @xmath3th roots of unity .
let @xmath0 be a prime of @xmath2 of norm @xmath37 .
for any such @xmath0 , @xmath38 and @xmath39 , one may associate a gauss sum @xmath40 modulo @xmath0 .
gauss sums have a rich arithmetic structure , but here we will only require that they are complex numbers that satisfy the relations below
. the interested reader may consult ( * ? ? ?
* section 2 ) for precise definitions .
in particular we note that the gauss sums we describe are generalizations of traditional finite field gauss sums , whose properties are detailed in , for example , ( * ? ? ?
* chapter 8) .
first we have @xmath41 where @xmath42 is the size of the residue field associated with @xmath43 .
moreover , if @xmath44 does not vanish modulo @xmath1 , then @xmath45 as we will most often deal with the situation when @xmath46 , we denote @xmath47 by @xmath48 .
we now review the chinta gunnells construction @xcite of @xmath0-parts of weyl group multiple dirichlet series .
the idea is to use averaging in a way analogous to the weyl character formula to define a rational function supported on @xmath12 that is invariant under a certain weyl group action .
we first define this action and then explain the relationship between the @xmath0-parts and the invariant function .
fix an @xmath5-tuple of nonnegative integers @xmath49 .
the tuple @xmath50 is a _ twisting parameter _ that determines a strongly dominant weight @xmath51 ( also referred to as the twisting parameter ) and a @xmath17-action on @xmath12 : @xmath52 in particular , when @xmath53 is a simple reflection , we have @xmath54 consider @xmath55 $ ] , the ring of laurent polynomials on @xmath12 . the ring @xmath56 consists of all expressions @xmath57 of the form @xmath58 with @xmath59 almost all zero .
multiplication in @xmath56 is defined by addition in @xmath12 : @xmath60 and we identify @xmath56 with @xmath61 $ ] via @xmath62 . our goal is to define a weyl group action on the field of fractions @xmath63 of @xmath56 .
first , define a change of variables action on @xmath56 by @xmath64 this action is essentially a reformulation of the standard action of @xmath17 on @xmath12 .
one can check that if @xmath65 is a monomial , then @xmath66 for @xmath67 , let @xmath68 , and consider the sublattice @xmath69 generated by the set @xmath70 .
define @xmath71 as the set of functions @xmath72 such that @xmath73 lies in the kernel of the map @xmath74 .
then we have the decomposition @xmath75 we are now ready to define the chinta gunnells action .
fix @xmath76 . for @xmath77 , define @xmath78 .
let @xmath79 be the normalized gauss sum @xmath80 for positive integers @xmath81 and @xmath82 , let @xmath83 be the remainder of @xmath81 upon division by @xmath82 .
also define @xmath84 if @xmath85 and @xmath86 otherwise .
define the following rational functions : @xmath87 then the simple reflection @xmath88 acts on @xmath89 by @xmath90 and this action satisfies the defining relations for @xmath17 ( * ? ? ?
* theorem 3.2 ) .
finally , we average this action applied to @xmath91 over the weyl group to obtain the invariant function whose numerator yields the @xmath0-parts .
first for @xmath92 , let @xmath93 be the number of @xmath94 in any reduced expression and put @xmath95 .
then define polynomials @xmath96 and @xmath97 .
the invariant function we seek is @xmath98 in particular , for @xmath99 the product @xmath100 is polynomial in the @xmath101 ( * ? ? ?
* theorem 3.5 ) . for each @xmath50 ,
the @xmath0-part is the polynomial @xmath102 . for details on how the @xmath0-parts determine the global series coefficients , see ( *
* section 4 ) .
we think of the @xmath0-parts as `` metaplectic '' symmetric functions .
indeed when @xmath9 , we recover a deformation of the weyl character formula from @xmath103 .
the main goal of this section is to describe the support of the @xmath0-parts @xmath102 .
the key tool will be an explicit set of recurrence relations on the coefficients of @xmath102 .
the invariance of @xmath104 under the @xmath17-action induces a recurrence relation on the coefficients of @xmath102 .
this recurrence relation is summarized in the following theorem , which extends ( * ? ? ?
* theorem 3.1 ) from @xmath4 simply laced and @xmath10 to all @xmath4 and @xmath1 .
[ thm : rec ] let @xmath105 and fix a simple reflection @xmath106 .
let @xmath107 , @xmath108 , @xmath109 , @xmath110 , and @xmath111 . then ,
if @xmath112 , we have @xmath113 put @xmath114 .
let @xmath115 and @xmath116 , with @xmath117 . for all @xmath92
, we have ( * ? ? ?
* lemma 3.4 ) @xmath118 it follows from @xmath119 that @xmath120 .
the @xmath17-invariance of @xmath104 yields @xmath121 and one checks that @xmath122 for @xmath77 , we calculate the @xmath123 coefficient in . to isolate the terms on the right - hand side that contribute , we define new functions
@xmath124 substituting into using , we obtain @xmath125|p|^{n - l+\delta(\lambda)}{{\bf x}}^{n\alpha+\sigma\lambda},\ ] ] where we have used that the change of variables action under @xmath126 takes @xmath123 to @xmath127 and that @xmath128 .
a straightforward calculation shows that the terms on the right - hand side that contribute to the @xmath123 coefficient are @xmath129 for convenience , we show the computation for the @xmath130 term in the case @xmath131 . for any @xmath132 ,
the monomial contribution from @xmath133 is @xmath134 .
we need only check that the exponent is @xmath135 when @xmath136 .
this requires simplifying @xmath137+[(n-(\delta(\sigma\bullet\lambda+\nu\alpha))_n ) \alpha ] $ ] .
the first term is @xmath138 .
thus it suffices to show that the second term is @xmath139 or equivalently that @xmath140 we have @xmath141 .
it follows that @xmath142 , and @xmath143 checking the contributions for @xmath144 and @xmath145 is similar . for the @xmath144 term
, one must show that @xmath146 when @xmath147 . for the @xmath145 term
, we need that @xmath148 when @xmath149 .
these statements are both clear from definition . collecting the @xmath123 coefficients and moving all terms of to the right - hand side
, we obtain a five term recurrence relation : @xmath150 where in we put @xmath151 .
we note that is a generalization of a relation in @xcite , which applied to the case @xmath8 and @xmath152 .
we next apply a second time , now with @xmath153 as the monomial on the left - hand side .
first , we calculate the contributions to the @xmath153 coefficient on the right - hand side .
we have @xmath154 again , when we collect the coefficients , now moving all terms to the left - hand side , we obtain a second five - term recurrence relation : @xmath155 where in we put @xmath156 .
note that in comparison with , we have multiplied each term by @xmath157 . adding and and simplifying , we obtain the @xmath1 recurrence relations and stated in the theorem . for @xmath158
, we note that when @xmath156 , we have @xmath159 similarly , when @xmath151 , we have @xmath160 we are now able to describe the support of @xmath102 . in particular , we show that @xmath102 is supported on a shifted weight polytope for a lowest weight representation .
recall that @xmath161 .
let @xmath162 be the convex hull of the points @xmath163 for @xmath92 .
more precisely , @xmath162 is the weight polytope for the irreducible representation of _ lowest _ weight @xmath164 , shifted by @xmath165 .
if @xmath166 is the set of dominant weights in the representation of _ highest _ weight @xmath165 , then all points of @xmath162 have the form @xmath167 where @xmath92 and @xmath168 .
our result is that the only nonzero coefficients of @xmath102 are those associated to the vertices of @xmath162 .
the following theorem extends ( * ? ? ?
* theorem 3.2 ) to all @xmath4 and all @xmath1 .
[ thm : support ] the support of @xmath102 is contained in @xmath162 .
we argue as in @xcite using the recurrence relation of theorem [ thm : rec ] .
we require two geometric lemmas .
note that @xmath162 is cut out by the inequalities @xmath169 this system of inequalities is redundant in the sense of the following lemma .
* lemma 3.4 ) [ lem : ineq ] let @xmath170 and let @xmath171 .
if @xmath172 , the inequalities @xmath173 and @xmath174 are equivalent . moreover , @xmath162 has bounded support .
* lemma 3.5 ) [ lem : ray ] let @xmath175 be a vertex of @xmath162 , and suppose @xmath176 .
then any lattice point of the form @xmath177 , where @xmath82 is a positive integer , lies outside of @xmath162 .
similarly , let @xmath178 and let @xmath179 .
then any point of the form @xmath180 , where @xmath82 is a positive integer , lies outside of @xmath162 . to avoid repetition ,
we summarize the key ideas here and refer the reader to @xcite for details . recall that @xmath105 is a polynomial .
the goal is to show that for @xmath181 we have @xmath182 .
equivalently we show that if @xmath183 for some @xmath92 and @xmath27 , then @xmath182 .
we induct on the length of @xmath184 .
if @xmath185 and @xmath135 violates the inequalities active at the origin , then clearly @xmath182 .
otherwise we would have polar terms , a contradiction since @xmath102 is polynomial .
now suppose that @xmath186 and that we have verified the inequalities at all vertices where @xmath187 , where @xmath188 .
it follows from lemma 3.3 and the induction hypothesis that it suffices to consider the case when @xmath189 for some @xmath190 and to prove @xmath182 if @xmath135 violates the inequality @xmath191 for the sake of contradiction , let @xmath192 and choose @xmath193 such that @xmath194 violates , but @xmath195 .
further assume that @xmath196 for all @xmath197 with @xmath198 .
in other words , @xmath194 is the final point of support on the ray @xmath199 . by lemma [ lem : ray ] ,
such a @xmath194 exists .
we apply theorem [ thm : rec ] with @xmath200 to @xmath201 , where @xmath201 is the first coefficient on the right - hand side of the relation .
then @xmath202 for some @xmath77 .
here we set @xmath203 and @xmath204 .
we have @xmath205 for all @xmath198 , and if @xmath206 then @xmath207 .
thus , the right - hand side yields @xmath208 regardless of the value of @xmath209 . in the case
@xmath210 , the left - hand side is @xmath211 otherwise , we have @xmath212 one checks that @xmath213 vanishes by the induction hypothesis , since @xmath214 implies @xmath215 violates the inequalities active at @xmath216 .
again by lemma [ lem : ray ] this implies @xmath217 and @xmath218 also vanish .
therefore , the left - hand side is identically zero and @xmath201 vanishes .
it follows that @xmath219 unless @xmath194 satisfies .
we now take a slight detour from our main goal to consider the support the modified @xmath0-parts @xmath220 .
as mentioned in the introduction , the @xmath221 coincide , up to a change of variables , with weyl group multiple dirichlet series associated to the rational function field over a finite field .
we expect that weyl group multiple dirichlet series associated to function fields over finite fields encode arithmetic data about the arithmetic of the defining curve , much like zeta functions .
theorem [ thm : gap ] has applications toward understanding the support of these series , which the author will explore in a forthcoming paper .
we include the result here as a complement to theorem [ thm : support ] .
[ thm : gap ] the support of @xmath221 lies outside the polytope @xmath162 .
figures [ fig : a2gap00 ] and [ fig : b2gap00 ] illustrate theorem [ thm : gap ] when @xmath222 and @xmath4 is @xmath223 or @xmath224 .
the plotted points represent the nonzero coefficients of @xmath225 , where @xmath226 corresponds to @xmath227 .
.49 .49 .49 .49 before giving a proof , we recall the following .
let @xmath228 be the set of positive roots made negative by @xmath184 .
one can show that @xmath229 .
if @xmath230 , then ( * ? ? ?
* section 5.6 ) @xmath231 in particular , if @xmath232 , then @xmath233 by definition @xmath234 .
thus , it is enough to show that the support of @xmath235 is contained in the cone determined by @xmath236 and shifted by @xmath163 .
we prove this claim by induction on @xmath93 .
when @xmath185 , the statement is clear .
now assume that @xmath232 .
we wish to show that @xmath237 is supported on the cone determined by @xmath238 and shifted by @xmath239 . by our assumption on the relative lengths of @xmath184 and @xmath240
, we have @xmath241 since @xmath242 for all @xmath184 , it follows from that @xmath243({{\bf x}})=j(w,\sigma_k{{\bf x}})[f|_\ell({{\bf x } } ) ] .
$ ] by ( * ? ? ?
* lemma 3.3 ) , up to sign we have @xmath244 where we have used that @xmath245 . from the above equality
, we see @xmath246({{\bf x}})\\ & = |p|^{n(\alpha_k)d(\alpha_k)}{{\bf x}}^{n(\alpha_k)\alpha_k}\left(j(w,\sigma_k{{\bf x}})[(1|_\ell w)|_\ell\sigma_k]({{\bf x}})\right)\\ & = |p|^{n(\alpha_k)d(\alpha_k)}{{\bf x}}^{n(\alpha_k)\alpha_k}\left([j(w,{{\bf x}})(1|_\ell w)]|_\ell \sigma_k\right)({{\bf x}}).\end{aligned}\ ] ] our inductive hypothesis implies that @xmath235 is supported on the cone defined by @xmath236 and shifted by @xmath163 .
thus , we can assume that @xmath247
where each @xmath248 is a monomial supported on @xmath236 shifted by @xmath163 .
using , we have @xmath249|_\ell \sigma_k\right)({{\bf x}})&=\sum [ p_{\beta,\ell , k}(x_k)+q_{\sigma_k\bullet \beta,\ell , k}(x_k)]f_\beta(\sigma_k{{\bf x}}),\end{aligned}\ ] ] where up to constants @xmath250 , @xmath251 by definition , we see that @xmath252 where @xmath253 and @xmath254 are rational functions supported on the ray determined by @xmath255 . after multiplying by @xmath256 ,
each of these two terms is supported on @xmath257 .
moreover , since the original cone had been shifted to the vertex @xmath258 , which corresponds to @xmath184 , the new cone is shifted to the vertex @xmath259 , which corresponds to a reflection under @xmath260 .
in this section we compare the methods of @xcite and @xcite to define @xmath0-parts of weyl group multiple dirichlet series .
more specifically , we compare these methods in the `` stable '' case @xmath152 . in this case , the only nonzero coefficients of @xmath102 are those attached to the vertices of the polytope @xmath162 @xcite .
we call these coefficients the stable coefficients . for the precise stability condition , see ( * ? ?
? * equation 20 ) .
when @xmath4 is type @xmath56 , it is enough to have @xmath261 . as in @xcite
, we caution the reader that we now make a slight change in notation : the element @xmath184 now corresponds to the coefficient of @xmath262 . this change is to aid the comparison of the sets @xmath263 and @xmath264 . for each @xmath165
, @xcite defines the stable coefficients @xmath265 where @xmath266 for some @xmath92 , and @xmath267 .
[ thm : stable ] assume that we are in the stable case @xmath268 .
let @xmath269 , and let @xmath105 be the @xmath0-part constructed via @xcite .
then @xmath270 in other words , the stable coefficients of @xcite and @xcite agree .
we follow @xcite and induct on @xmath93 .
assume that @xmath271 .
if @xmath185 , we have an empty product on the right hand side of and the statement holds trivially .
suppose that @xmath272 on all @xmath273 with @xmath274 .
let @xmath275 , with @xmath276 .
in other words , we assume @xmath277 and @xmath278 . applying or with @xmath201 on the right - hand side ,
the outer terms vanish by lemma [ lem : ray ] . we have @xmath279 if @xmath280 and @xmath281 otherwise . on the other hand , it follows from and our induction hypothesis that @xmath282 it remains to show @xmath283 when @xmath284 , we have @xmath285 . otherwise , we use and to rewrite @xmath286 from
, it suffices to show that @xmath287 . by definition @xmath288
also @xmath289 this implies the result .
in this section we address the extent to which @xmath290 is completely determined by theorem [ thm : rec ] .
theorem [ thm : unstab ] states that the set of all such @xmath102 form a @xmath291-vector space with dimension at most the number of regular dominant weights in the representation @xmath292 of highest weight @xmath165 .
this result extends ( * ? ? ?
* theorem 5.7 ) to all @xmath4 and all @xmath1 , and shows that when @xmath293 , the @xmath0-part @xmath102 is completely determined by theorem [ thm : rec ] after setting @xmath271 .
our argument follows @xcite .
the idea is to consider @xmath184 orbits of the coefficients .
recall that by theorem [ thm : support ] , the support of @xmath105 consists of all @xmath294 such that @xmath92 and @xmath168 . for any @xmath168 , define @xmath295 as the @xmath17-orbit of the coefficient @xmath296 under the @xmath297 action .
let @xmath298 be the set of all such orbits .
there is a natural partial order on @xmath299 given by the poset relation on the weights : we say @xmath300 if and only if @xmath301 . under this identification of @xmath302 with @xmath303 ,
the condition @xmath301 becomes @xmath304 .
[ thm : unstab ] let @xmath305 .
define @xmath306 to be the set of all regular dominant weights in the representation of highest weight @xmath165 .
suppose that the coefficients @xmath296 of @xmath102 are known for all @xmath307 .
then @xmath102 is completely determined by the relations or after setting @xmath271 .
we induct on the poset @xmath308 .
first , we know all the points in the orbit @xmath309 by theorem [ thm : stable ] .
now fix @xmath168 such that @xmath310 , and assume that we have determined the coefficients attached to all orbits @xmath311 with @xmath312 .
we consider two cases : @xmath302 regular or not .
assume that @xmath302 is regular , and let @xmath313 .
by assumption , the value of the coefficient @xmath213 associated with @xmath302 is known .
we must show that we can determine @xmath314 for all @xmath315 .
any such @xmath316 is of the form @xmath317 for @xmath92 and thus can be obtained from successive application of simple reflections @xmath94 . from relation or , when we apply @xmath94 to @xmath213 , the outer term on the left - hand side is either @xmath318 or @xmath319 , where @xmath203 and @xmath320 . in both cases , ( * ? ? ?
* lemma 5.3 ) implies that these terms come from a previously determined orbit @xmath321 with @xmath322 . under the action
@xmath323 we obtain the outer term on the right - hand side , so this term also belongs to @xmath321 and hence is previously determined .
since we know three out of four terms of the relation , @xmath314 is determined .
now , assume that @xmath302 is not regular .
since @xmath324 , there exists a simple reflection @xmath94 such that @xmath213 is taken to itself under relation or . by ( * ? ? ?
* lemma 5.5 ) , all other @xmath325 involved in the recurrence are predetermined ; therefore we know @xmath213
. we may now successively apply or to determine the remaining coefficients @xmath167 in the orbit @xmath326 . | we study the structure of @xmath0-parts of weyl group multiple dirichlet series .
in particular , we extend results of chinta , friedberg , and gunnells @xcite and show , in the stable case , that the @xmath0-parts of chinta and gunnells @xcite agree with those constructed using the crystal graph technique of brubaker , bump , and friedberg @xcite . in this vein , we give an explicit recurrence relation on the coefficients of the @xmath0-parts , which allows us to describe the support of the @xmath0-parts and address the extent to which they are uniquely determined . |
one - dimensional ( 1d ) quantum spin systems have received continued attention as paradigms for strongly correlated systems , because miscellaneous and even exotic
phases can be realized within simple model hamiltonians . by way of example
the exactly solvable spin-1/2 antiferromagnetic ( afm ) heisenberg chain is known to be gapless , while for integer spin a gap exists between the ground state and the first excited state , as conjectured first by haldane.@xcite especially for the spin-1 chain , the haldane gap was confirmed experimentally,@xcite and the dynamical spin structure factor has been observed by inelastic neutron scattering , e.g. , on ni(c@xmath1h@xmath2n@xmath1)@xmath1no@xmath1clo@xmath3.@xcite affleck , lieb , kennedy , and tasaki ( aklt ) proposed a exactly solvable model that offers valuable clues to the physics of the spin-1 heisenberg chain.@xcite the so - called aklt state [ cf .
1 ( a ) below ] successfully describes the ground state of the haldane phase.@xcite also for the the spin-1 @xmath0 model , the ground - state phase diagram has been determined even if a single - ion anisotropy is added @xcite e.g .
, by the lanczos exact diagonalization ( ed ) technique based on the level spectroscopy method.@xcite currently , quantum integer - spin chains have attracted extraordinary interest from a topological point of view .
the gapped ground states in the haldane phase can be classified by the projective representations of the underlying symmetry group.@xcite the odd haldane ( oh ) phase in odd - integer spin chains with two half - integer edge spins is a symmetry - protected topological ( spt ) phase , because the odd-@xmath4 aklt state can not be adiabatically connected to another trivial state without undergoing a phase transition . on the other hand ,
the even haldane ( eh ) state in the even - integer spin systems with integer edge spins@xcite is a trivial state , since the even-@xmath4 aklt state is adiabatically connected to a trivial state without a bulk phase transition.@xcite interestingly , a hidden spt phase analogous to the oh phase was discovered in the extended bose - hubbard model ( ebhm ) with longer - range repulsions.@xcite this haldane insulator ( hi ) phase , embedded between the mott insulator ( mi ) and the density wave ( dw ) phases in the intermediate coupling regime , exhibits the characteristic degeneracy of the entanglement spectrum in the haldane phase.@xcite the excitation gaps at the quantum phase transition lines depend on their universality classes.@xcite beyond that , the dynamical density structure factor @xmath5 significantly differs in the mi , dw , and hi states.@xcite on the basis of our recent ebhm study,@xcite in the present work , we investigate the topological properties of the odd haldane phase in the anisotropic spin-1 @xmath0 chain which , as we will show , can be taken as an effective model for the ebhm . using the density matrix renormalization group ( dmrg )
technique,@xcite first we determine the phase boundaries by exploiting the central charge . in order to confirm the closing of the excitation gap at the trivial - nontrivial phase transition points , we simulate both the spin and neutral gaps .
we furthermore demonstrate the degeneracy of entanglement levels in the oh phase with both periodic ( p ) and open ( o ) boundary conditions ( bc ) [ for the anisotropic spin-1 @xmath0 chain it is well known how the edge spins should be treated in the latter case ] . in order to experimentally detect the topological hi phase in the ebhm
, various dynamical quantities have been proposed.@xcite here we will examine the dynamical spin structure factor @xmath6 for the spin-1 model by means of the dynamical dmrg ( ddmrg ) technique.@xcite we will demonstrate that the intensity maximum in @xmath6 features a gapped dispersion in the non - trivial haldane phase as obtained for @xmath5 in the ebhm .
since this quantity is directly accessible by inelastic neutron scattering , significant differences in @xmath6 could be used to detect the various gapped phases .
this paper is organized as follows . in the next section
we establish the anisotropic spin-1 @xmath0 model and the corresponding ebhm .
the physical quantities of interests are introduced in sec .
[ physical - quantities ] .
large - scale ( d)dmrg results for the anisotropic spin-1 @xmath0 chain will be presented and discussed in sec .
[ dmrg - results ] .
section [ summary ] contains a brief summary and our main conclusions .
in this section we introduce the anisotropic spin-1 @xmath0 model and get back to its established ground - state phase properties .
we then define the extended bose - hubbard model and point out the correspondences with an effective spin-1 @xmath0 model .
the hamiltonian of the 1d spin-1 @xmath0 model with on - site anisotropy is given by @xmath7 \nonumber \\ & & + d\sum_j ( \hat{s}_j^z)^2\ , \label{hamil}\end{aligned}\ ] ] where @xmath8 denotes a spin-@xmath9 operator .
the parameter @xmath10 represents the uniaxial single - ion anisotropy .
the ground - state phase diagram of the model exhibits various gapful and gapless phases , namely , following the conventional notations , the haldane phase , the large-@xmath10 phase , two @xmath11 phases , the ferromagnetic phase , and the nel phase.@xcite according to this different types of phase transitions occur between these phases : ( i ) a gapful - gapful gaussian phase transition takes place between the large-@xmath10 phase and the haldane phase with the central charge @xmath12 , ( ii ) the haldane - nel transition appears to be of the ising universality class with @xmath13 , and ( iii ) a gapless - gapful berezinskii - kosterlitz - thouless ( bkt ) transition emerges between the @xmath11 phase and the haldane or large-@xmath10 phase . in what follows
we restrict ourselves to the parameter region where @xmath14 and @xmath15 . following the notation by kjll _
et al._,@xcite we use the termini eh , oh , and afm phases instead of the large-@xmath10 , haldane , and nel phases , respectively .
the lattice inversion symmetry , which protects the spt state of the haldane phase , can be broken by adding a perturbation to the hamiltonian ( [ hamil ] ) : @xmath16\ , .
\label{pert - term}\end{aligned}\ ] ] any finite @xmath17 immediately lifts the characteristic degeneracy of the lowest entanglement level in the haldane phase.@xcite as we will see later , thereby the eh - oh quantum phase transition also disappears . in 2006 , dalla torre _
et al._@xcite discovered the hi phase in the 1d extended bose - hubbard model with longer - range repulsions .
the hi phase features the properties of the oh phase in the spin-1 model ( [ hamil ] ) .
the ebhm hamiltonian reads @xmath18 where @xmath19 ( @xmath20 ) creates ( annihilates ) a boson at lattice site @xmath21 , and @xmath22 is the corresponding boson number operator .
the nearest - neighbor boson transfer amplitude is given by @xmath23 and @xmath24 ( @xmath25 ) parametrizes the on - site ( nearest - neighbor ) particle repulsion . assuming that the site occupation is restricted to @xmath26 , 1 , or 2 , with @xmath27 for a mean boson filling factor @xmath28
, the system can be mapped onto an effective spin-1 hamiltonian , @xmath29 with the replacements @xmath30 , @xmath31 , and @xmath32 in eq . .
@xmath33 contains further terms which breaks the particle - hole symmetry of @xmath34 [ see eq . ( a1 ) of ref .
[ ] for the explicit form of @xmath33 ]
. the ebhm exhibits three insulating phases , where the nontrivial hi phase appears in between the mi and dw phases for intermediate - couplings .
the mi , hi , and dw phases of the ebhm correspond to the eh , oh , and afm phases of the spin-1 model ( [ hamil ] ) , respectively .
in this section we assort the quantities that can be used to characterize the different phases and phase transitions in the spin-1 model ( [ hamil ] ) and accordingly in the ebhm .
we furthermore explain how the quantities can be simulated using the dmrg technique . chain .
each of the two @xmath36 spins connected by an ellipse form a singlet ( @xmath37)(@xmath38 ) .
the two free edge - spins cause the fourfold degeneracy of the ground - state energy .
( b ) to simulate the aklt state within obc dmrg , the free edge spins have to be excluded from the system ( dashed circles ) .
( c ) with pbc the aklt state can be simulated without any changes , so that the lowest entanglement level exhibits a fourfold degeneracy according to the two edge spins .
, title="fig : " ] + chain .
each of the two @xmath36 spins connected by an ellipse form a singlet ( @xmath37)(@xmath38 ) .
the two free edge - spins cause the fourfold degeneracy of the ground - state energy .
( b ) to simulate the aklt state within obc dmrg , the free edge spins have to be excluded from the system ( dashed circles ) .
( c ) with pbc the aklt state can be simulated without any changes , so that the lowest entanglement level exhibits a fourfold degeneracy according to the two edge spins .
, title="fig : " ] after li and haldane s proposal @xcite to characterize topological phases by the entanglement spectrum this has become one of the most powerful tools to investigate the spt state . dividing a system with @xmath39 sites into two subblocks and considering the reduced density matrix @xmath40 $ ] of a subblock of arbitrarily length @xmath41
, the entanglement spectrum @xmath42 is obtained from the weights @xmath43 of the reduced density matrix @xmath44 by @xmath45 the entanglement spectrum of a subblock with @xmath46 sites can be obtained for obc and pbc as sketched in figs .
[ haldane - state](b ) and [ haldane - state](c ) , respectively .
thereby the artificial edges give rise to the characteristic degeneracy of the lowest entanglement level in the nontrivial aklt state [ displayed in fig .
[ haldane - state](a ) ] , where the degree of degeneracy depends on the boundary conditions . to determine the entanglement spectrum in the oh phase with obc , a well - known trick is in use :
one simulates a system without free edge spins by replacing the edge sites with @xmath36 , as shown in fig .
[ haldane - state](b ) .
one then expects a doubly degenerate lowest entanglement level in the oh phase .
for pbc , on the other hand , for the same finite system , a fourfold degeneracy is expected due to two free @xmath36 spins [ see fig .
[ haldane - state](c ) ] , just as for the hi phase in the ebhm.@xcite the entanglement analysis provides also valuable information about the criticality of the system .
adding up the @xmath43 during the simulation , we have direct access to the von neumann entropy @xmath47 $ ] . from conformal field
theory,@xcite it follows that in the case of a periodic system the von neumann entropy takes the form @xmath48 + s_1\ , , \end{aligned}\ ] ] where @xmath49 is a non - universal constant . since the most precise data of @xmath50 are obtained when the length @xmath41 of the subblock equals half the system size @xmath39 , the relation @xcite @xmath51}{\ln[\cos(\pi / l ) ] } \label{cstar}\end{aligned}\ ] ] is much better suited for determining the central charge than directly using the above expression for @xmath50 . for the ebhm
the phase boundaries can be assigned very effectively using the ( numerically determined ) central charge @xmath52 , because the system becomes critical only at the mi - hi ( hi - dw ) transition points where @xmath12 ( @xmath13 ) , and there @xmath52 shows pronounced peaks.@xcite hence we adopt this method for the spin-1 model as well to pinpoint the eh - oh and oh - afm transition points .
monitoring various excitation gaps for the ebhm , significant features have been found at the mi - hi and hi - dw transition points.@xcite for example , the single - particle gap , @xmath53 is finite in all three insulating phases , except for the mi - hi transition point . by contrast , the neutral gap , @xmath54 closes at both the mi - hi and hi - dw transitions . in eqs . and
, @xmath55 and @xmath56 denote the energies of the ground state and first excited state of the @xmath57-particle system for the ebhm , respectively .
since adding ( removing ) a particle in the ebhm corresponds to raising ( lowering ) the spin @xmath58 projection in a pseudospin model , we consider for the spin-1 @xmath0 model the spin gap , @xmath59 which likewise might be finite in all three phases , except for the eh - oh transition point . as for the ebhm , the neutral gap in the spin-1 model ( [ hamil ] ) can be defined as @xmath60 where @xmath61 and @xmath62 denote the ground - state and first excited energies within the subspace @xmath63 , respectively . by analogy to the behavior of the neutral gap in the ebhm , @xmath64 should vanish at the eh - oh and oh - afm transition points for the spin-1 chain model . simulating the dynamical spin structure factor by ddmrg is of particular importance
since it might be directly compared with inelastic neutron scattering experiments , e.g. , on ni(c@xmath1h@xmath2n@xmath1)@xmath1no@xmath1clo@xmath3.@xcite its @xmath65-component is defined by @xmath66 where @xmath67 and @xmath68 denote the
ground state and @xmath69th excited state , respectively .
the corresponding excitation energy is @xmath70 . for @xmath71 ,
i.e. , for the isotropic heisenberg or @xmath0 fix points of ( [ hamil ] ) , @xmath6 was extensively studied by ed @xcite and time - dependent dmrg @xcite techniques .
that is , the behavior of @xmath6 in the haldane phase is well known , albeit numerical results for the eh and afm states are rare .
taking into account the relation @xmath72 for the pseudospin in the effective model @xmath73 , one expects that the spin structure factor @xmath6 corresponds to the dynamical density structure factor @xmath5 in the ebhm , which exhibits different behavior in the three insulating phases.@xcite
in this section we present our numerical ( d)dmrg results for the spin-1 @xmath0 model with and without single - ion anisotropy .
we first determine the phase boundaries and then analyze the behavior of the excitation gaps at the transitions between the nontrivial and trivial phases .
furthermore , we discuss the entanglement spectra of an odd haldane phase . finally , we simulate the dynamic spin structure factor and compare it with the dynamical density response in the ebhm . in the numerics we keep up to @xmath74 density - matrix states for the static dmrg runs , so that the discarded weight is typically smaller than @xmath75 . for
the ddmrg simulations we take @xmath76 , examining the ground state along the first five dmrg sweeps , and then use @xmath77 states computing dynamical properties . as obtained by dmrg for the spin-1 @xmath0 model with @xmath71 and pbc .
the oh - afm transition can be assigned to @xmath78 with @xmath13 , where a pronounced peak appears [ see also the magnifying inset which shows @xmath79 close to the transition point ] . ]
let us first discuss the spin-1 model ( [ hamil ] ) with @xmath71 .
in this case it is known that a bkt transition occurs at @xmath80 between the @xmath11 and oh phases.@xcite at @xmath14 , only an oh - afm transition takes place , where @xmath13 is expected .
figure [ cstar - d0 ] shows the central charge @xmath79 , computed from eq . .
if @xmath81 is raised at fixed system size , the maximum in @xmath79 sharpens at the oh - afm transition point @xmath82 , and we deduce @xmath83 .
the other critical point @xmath84 with @xmath85 approximates the recent infinite - system dmrg result @xmath86@xcite very well already for @xmath87 .
the agreement becomes perfect if we increase the system size : @xmath88 with @xmath89 for @xmath90 .
note that @xmath79 stays equal to one in a relatively wide region ( from @xmath91 to @xmath92 for @xmath90 ) , indicating the bkt transition between the @xmath11 and the oh phases at @xmath91 with @xmath12 .
for @xmath93 near the oh - afm transition .
panel ( b ) shows that the ising transition point @xmath94 obtained from @xmath79 can be linearly extrapolated to the thermodynamic limit . ] to relate our numerical results to previous ones we include an on - site anisotropy @xmath10 and compute @xmath79 in the vicinity of the oh - afm transition . for @xmath95 the central charge @xmath79 at fixed system size
@xmath39 develops again a pronounced maximum at the oh - afm transition point [ see fig .
[ cstar - d0.5](a ) ] .
the deduced transition point @xmath96 is readily extrapolated to the thermodynamic limit [ fig .
[ cstar - d0.5](b ) ] , yielding @xmath97 , which is in reasonable agreement with the ed result @xmath98 obtained from systems with @xmath39 up to 16@xcite and confirms recent dmrg data @xmath99.@xcite -dependence of the two lowest energy eigenvalues at @xmath100 , using tbc and @xmath87 .
obviously the energies of the oh state ( squares ) and the eh state ( circles ) cross at the eh - oh transition point ( dashed line ) .
( b ) : critical points @xmath101 obtained by the level spectroscopy technique ( stars ) [ via the central charge ( pluses ) ] as obtained in panel ( a ) [ ( c ) ] versus the inverse of the system size squared at @xmath100 for @xmath39 up to 128 .
( c ) : central charge @xmath52 of the 1d spin-1 @xmath0 model ( [ hamil ] ) with @xmath100 , indicating the eh - oh ( oh - afm ) transition point with @xmath12 ( @xmath13 ) .
the solid line denotes the eh - oh transition extracted from panels ( a ) and ( b ) , which is in accordance with the position of the maximum in @xmath79 . turning on a perturbation @xmath102 that breaks the lattice - inversion symmetry , the central charge
@xmath79 ( filled symbols ) becomes zero for large enough system sizes ( @xmath103 ) .
, title="fig : " ] -dependence of the two lowest energy eigenvalues at @xmath100 , using tbc and @xmath87 .
obviously the energies of the oh state ( squares ) and the eh state ( circles ) cross at the eh - oh transition point ( dashed line ) .
( b ) : critical points @xmath101 obtained by the level spectroscopy technique ( stars ) [ via the central charge ( pluses ) ] as obtained in panel ( a ) [ ( c ) ] versus the inverse of the system size squared at @xmath100 for @xmath39 up to 128 .
( c ) : central charge @xmath52 of the 1d spin-1 @xmath0 model ( [ hamil ] ) with @xmath100 , indicating the eh - oh ( oh - afm ) transition point with @xmath12 ( @xmath13 ) .
the solid line denotes the eh - oh transition extracted from panels ( a ) and ( b ) , which is in accordance with the position of the maximum in @xmath79 . turning on a perturbation @xmath102 that breaks the lattice - inversion symmetry , the central charge @xmath79 ( filled symbols ) becomes zero for large enough system sizes ( @xmath103 ) .
, title="fig : " ] we now turn to the case @xmath15 . in previous works @xcite a gaussian transition between the eh and oh phases has been found by employing the level spectroscopy technique to ed results obtained for small systems . applying the twisted boundary conditions ( tbc ) , @xmath104 , @xmath105 , and @xmath106 within dmrg , the two lowest energy levels can be simulated accurately for much larger system sizes than accessible to ed .
figure [ twist - cstar - d1.5](a ) demonstrates that the two lowest energies assigned to the eh and oh states cross at @xmath107 by increasing @xmath108 at fixed @xmath100 for @xmath87 ( and tbc ) .
the critical points @xmath109 can be systematically extrapolated to the thermodynamic limit by a linear fit , as indicated in fig .
[ twist - cstar - d1.5](b ) .
for @xmath110 we obtain @xmath111 .
alternatively , the eh - oh transition points can be extracted from the central charge @xmath79 if compared with the field theoretical prediction @xmath12 .
this is demonstrated in fig .
[ twist - cstar - d1.5](c ) . here
the maxima of @xmath79 can also be extrapolated to the thermodynamic limit [ see fig . [ twist - cstar - d1.5](b ) ] , where transition point is in excellent accord with the ones via level spectroscopy in figs . [ twist - cstar - d1.5](a ) and [ twist - cstar - d1.5](b ) .
we note that also the oh - afm transition can be reliably determined from the peak at @xmath112 . in the following
we analyze the signatures of the topological oh phase and of the transition between the trivial and nontrivial topological states for the model in close analogy to the ebhm.@xcite to this end , we simulate the excitation gaps and the entanglement spectra . ( squares ) and neutral gap @xmath113 ( open circles ) as a function of @xmath108 for @xmath114 [ panel ( a ) ] and @xmath100 [ panel ( b ) ] .
the filled squares in panel ( b ) give @xmath115 with a finite inversion - symmetry breaking perturbation @xmath116 [ see eq . ] .
, title="fig : " ] ( squares ) and neutral gap @xmath113 ( open circles ) as a function of @xmath108 for @xmath114 [ panel ( a ) ] and @xmath100 [ panel ( b ) ] .
the filled squares in panel ( b ) give @xmath115 with a finite inversion - symmetry breaking perturbation @xmath116 [ see eq . ] .
, title="fig : " ] so far the excitation gaps of have been studied mostly at the isotropic heisenberg point with respect to the magnitude of the haldane gap . at the trivial - nontrivial phase transition
points the excitation gaps should close , as demonstrated , e.g. , for the ebhm.@xcite here we compute the spin and neutral excitation gaps as defined in sec .
[ sec : gap ] instead of calculating the simple first excitation gap .
thereby , we adopt pbc instead of obc within dmrg , avoiding the use of edge spins , which have to be adapted according to the considered parameter region .
figure [ gaps](a ) shows first the excitation gaps at @xmath71 . upon increasing @xmath108 , the gaps open exponentially , reflecting the bkt transition at @xmath117 .
@xmath113 and @xmath115 cross each other exactly at the heisenberg point , @xmath118 , where @xmath119 ( see the discussion about the system - size dependence of the excitation gaps and the magnitude of the haldane gap for the spin-1 heisenberg model in the appendix ) . at the oh - afm transition ( @xmath120 ) ,
@xmath64 closes linearly because the transition belongs to the ising universality class , while @xmath121 remains finite . for @xmath100
[ see fig .
[ gaps](b ) ] , the eh - oh transition occurs at @xmath122 , where both spin and neutral gaps vanish . increasing @xmath108 above @xmath123 , only @xmath64 closes at the ising transition point @xmath124 , just as in the case of @xmath114 [ compare figs .
[ gaps](a ) and [ gaps](b ) ] .
if we turn on the perturbation @xmath102 [ see eq .
] , which breaks the lattice - inversion symmetry explicitly , the eh - oh transition disappears , so that @xmath121 stays finite for @xmath116 , as shown in fig .
[ gaps](b ) . thereby
, owing to the loss of the criticality at the eh - oh transition , @xmath79 converges to zero for large enough @xmath39 , as demonstrated in fig .
[ twist - cstar - d1.5](c ) . comparing the behavior of the excitation gaps with those of the ebhm,@xcite
one sees that the spin ( neutral ) gap in the spin-1 model takes the role of the single - particle ( neutral ) gap in the ebhm . of the 1d spin-1 @xmath0 model ( [ hamil ] ) with @xmath125 .
the fourfold degeneracy of the lowest entanglement level can be taken as an indication of a nontrivial haldane state in the case of dmrg simulations with pbc .
as the system size increases the degeneracy appears in the whole hi phase ; compare data for @xmath126 [ ( a ) ] with those for @xmath127 [ ( b ) ] . using obc and taking spins ( @xmath36 ) at the edges into account ,
an almost perfect double degeneracy is obtained for the oh phase even for small systems with @xmath90 , as demonstrated in ( c ) .
, title="fig : " ] of the 1d spin-1 @xmath0 model ( [ hamil ] ) with @xmath125 .
the fourfold degeneracy of the lowest entanglement level can be taken as an indication of a nontrivial haldane state in the case of dmrg simulations with pbc .
as the system size increases the degeneracy appears in the whole hi phase ; compare data for @xmath126 [ ( a ) ] with those for @xmath127 [ ( b ) ] . using obc and taking spins ( @xmath36 ) at the edges into account ,
an almost perfect double degeneracy is obtained for the oh phase even for small systems with @xmath90 , as demonstrated in ( c ) .
, title="fig : " ] of the 1d spin-1 @xmath0 model ( [ hamil ] ) with @xmath125 .
the fourfold degeneracy of the lowest entanglement level can be taken as an indication of a nontrivial haldane state in the case of dmrg simulations with pbc .
as the system size increases the degeneracy appears in the whole hi phase ; compare data for @xmath126 [ ( a ) ] with those for @xmath127 [ ( b ) ] . using obc and taking spins ( @xmath36 ) at the edges into account ,
an almost perfect double degeneracy is obtained for the oh phase even for small systems with @xmath90 , as demonstrated in ( c ) .
, title="fig : " ] let us now analyze the entanglement properties of the topological states for intermediate single - ion anisotropy ( @xmath100 ) , where both the eh - oh and oh - afm transitions exist . here
et al._@xcite showed that the spt state in the oh phase has a twofold degenerate lowest entanglement level for the quantum spin chain model .
the infinite - time evolving block decimation procedure used by those authors , gives the entanglement spectra data directly in the thermodynamic limit . in the following
we show that when simulating the model for a finite system by conventional dmrg , this characteristic degeneracy of the oh phase can also be obtained , but the degree of the degeneracy depends on the boundary conditions .
figure [ es - d1.5 ] presents the entanglement spectrum @xmath42 for the anisotropic spin-1 @xmath0 model with @xmath100 . for a small system ( @xmath126 ) with pbc [ fig .
[ es - d1.5](a ) ] the lowest entanglement level is fourfold degenerate only deep inside the oh phase .
this reflects the possession of the two edges for the subblock @xmath128 . increasing the system size
this degeneracy is observed for a larger region of the oh phase , as demonstrated by fig .
[ es - d1.5](b ) for @xmath90 , but close to the eh - oh transition point the lowest entanglement level is still non - degenerate . to overcome this drawback we apply obc with half - spin edges in the oh phase [ cf .
[ haldane - state](b ) ] . the same procedure has been used to estimate the magnitude of the haldane gap at the isotropic heisenberg point .
figure [ es - d1.5](c ) gives @xmath42 for @xmath90 and obc , pointing out the twofold degeneracy of the lowest level in the nontrivial phase and its non - degeneracy anywhere else .
the degeneracy is clearly caused by the single edge spin of subblock @xmath128 . if an inversion - symmetry - breaking term is added to the spin-1 chain ( [ hamil ] ) with @xmath125 , where @xmath116 ( left panels ) and @xmath129 ( right panels ) .
data obtained by dmrg with pbc ( upper panels ) and obc with half - spin edges ( lower panels ) . , title="fig : " ] if an inversion - symmetry - breaking term is added to the spin-1 chain ( [ hamil ] ) with @xmath125 , where @xmath116 ( left panels ) and @xmath129 ( right panels ) .
data obtained by dmrg with pbc ( upper panels ) and obc with half - spin edges ( lower panels ) .
, title="fig : " ] recently it has been demonstrated for quantum spin chains@xcite and the ebhm@xcite that the degeneracy of the lowest entanglement level in the oh phase might be lifted by turning on an inversion - symmetry - breaking term , such as .
figure [ es - d1.5-perturb](a ) [ figure [ es - d1.5-perturb](c ) ] exemplifies that the fourfold [ twofold ] degeneracy with pbc [ obc ] indeed dissolves for any finite @xmath17 .
thereby , the gap between the lowest levels becomes larger as @xmath17 increases , see fig .
[ es - d1.5-perturb](b ) [ fig .
[ es - d1.5-perturb](d ) ] for pbc [ obc ] .
obviously inversion symmetry protects the haldane phase . for @xmath130 ( a ) , @xmath131 ( b ) , and @xmath132 ( c ) close to the eh - oh - afm multicritical point with @xmath133 and obc , showing the disappearance of the double degeneracy of the lowest entanglement levels .
, title="fig : " ] for @xmath130 ( a ) , @xmath131 ( b ) , and @xmath132 ( c ) close to the eh - oh - afm multicritical point with @xmath133 and obc , showing the disappearance of the double degeneracy of the lowest entanglement levels .
, title="fig : " ] for @xmath130 ( a ) , @xmath131 ( b ) , and @xmath132 ( c ) close to the eh - oh - afm multicritical point with @xmath133 and obc , showing the disappearance of the double degeneracy of the lowest entanglement levels .
, title="fig : " ] raising the ratios @xmath134 and @xmath108 , the eh - oh and oh - afm critical lines merge at the multicritical point ( @xmath135 , @xmath136 ) . above this point ,
i.e. , for @xmath137 and @xmath138 , a direct eh - afm transition is expected to occur , as pointed out by den nijs and rommeles.@xcite this has been confirmed numerically by ed , yielding ( @xmath139 , @xmath140 ) @xmath141 ( 3.2 , 2.9).@xcite it is challenging to determine this multicritical point more precisely , but the entanglement analysis outlined above seems to be a powerful tool . obviously , for fixed values @xmath142 , the lowest entanglement level is non - degenerate for the whole parameter regime of @xmath108 including the eh - afm transition point , while for @xmath143 the lowest entanglement level should be degenerate for a very narrow but finite parameter region of @xmath108 .
in fact , the double degeneracy can still be observed at @xmath130 by large - scale dmrg simulations with @xmath133 and obc [ see fig .
[ tricritical](a ) ] .
if @xmath134 is increased slightly , the degeneracy is lifted for @xmath133 [ see fig .
[ tricritical](b ) for @xmath144 , but from the results presented we can not derive a definite conclusion about what happens for @xmath110 .
figure [ tricritical](c ) indicates that degeneracy disappears already at @xmath145 . in this way the ed results regarding the existence of the multicritical point is corroborated by our more precise entanglement spectra analysis , yielding ( @xmath135 , @xmath136)=(@xmath146 , @xmath147 ) .
certainly it is of great interest to look at the behavior of the central charge at the multicritical point . here
the central charge might be @xmath148 because bkt- and ising - transition lines merge .
figure [ cstar - multi - eh - afm](a ) displays the numerically obtained central charge @xmath79 for fixed values of @xmath134 in the vicinity of the multicritical point .
we see that @xmath79 is always smaller than 3/2 and decreases with increasing system size @xmath39 .
unfortunately , the system - size dependence of @xmath79 is much stronger than those , e.g. , at @xmath100 , in fig .
[ twist - cstar - d1.5 ] ; so it turns out that even @xmath90 is not large enough to determine the value of the central charge precisely .
maybe the use of the infinite - system dmrg @xcite can resolve this problem . increasing @xmath134 further , a quantum phase transition occurs between eh and afm phases .
a discontinuous staggered magnetization suggests that this transition is of first order.@xcite quite recently , this was corroborated by analyzing the energy level crossing.@xcite the numerically determined central charge @xmath79 at @xmath149 yields a further signature of the first - order transition [ see fig .
[ cstar - multi - eh - afm](b ) ] . for small system sizes ( @xmath87 )
, @xmath52 shows a peak at @xmath150 , in accord with the eh - afm transition point in ref .
[ ] . with increasing the system size @xmath39 , @xmath52 decreases drastically and becomes already zero for @xmath90 , which confirms the results of former studies.@xcite of the 1d spin-1 @xmath0 model ( [ hamil ] ) near the multicritical point @xmath136 ( a ) and across the first - order eh - afm transition for @xmath149 ( b ) .
the dashed line in panel ( b ) denotes the eh - afm transition point @xmath151 according to ref .
title="fig : " ] of the 1d spin-1 @xmath0 model ( [ hamil ] ) near the multicritical point @xmath136 ( a ) and across the first - order eh - afm transition for @xmath149 ( b ) .
the dashed line in panel ( b ) denotes the eh - afm transition point @xmath151 according to ref .
[ ] . , title="fig : " ] figure [ pd ] displays the dmrg ground - state phase diagram of the spin-1 @xmath0 model with single - ion anisotropy .
the eh - oh and oh - afm phase boundaries can been derived from central charge @xmath52 , as explained above : again we obtain a very good agreement with former ed and dmrg data.@xcite most notably , the nontrivial oh phase appears in between the trivial eh and afm phases , just as the topological hi phase develops between the mi and dw phases in the ebhm .
therefore , we have included in fig .
[ pd ] , the phase boundaries of the mi - hi and hi - dw transitions for the ebhm with @xmath152 ( taken from ref . [ ] ) .
qualitatively , the phase diagram of the spin-1 model looks quite similar to those of the ebhm , except for the existence of the superfluid ( sf ) phase in the ebhm ( not shown ) .
quantitatively , the topological phase of the ebhm captures a larger region in parameter space than the oh phase however .
this might be caused by the particle - hole symmetry - breaking term @xmath153 in eq . .
model with single - ion anisotropy . shown
are the even haldane ( eh ) , odd haldane ( oh ) , and antiferromagnetic ( afm ) phases . the eh - oh ( squares ) and oh - afm ( circles ) transition points are determined from the central charge @xmath12 and @xmath13 , respectively , which was extracted from the von neumann entropy via eq . . the eh - oh transition line
was confirmed by a careful finite - size scaling of the two low - lying energy levels with tbc .
the filled diamond gives the eh - oh - afm multicritical point determined from the entanglement analysis .
error bars are smaller than symbols .
the dashed ( dotted ) line denotes the mi - hi ( hi - dw ) transition in the ebhm with @xmath152 bosons per site ( taken from ref . [ ] ) . ]
let us finally discuss the spin dynamical properties of the spin-1 @xmath0 model .
figure [ skw ] reveals our ddmrg results for @xmath6 obtained for the spin-1 model ( with anisotropy @xmath100 ) inside the three insulating phases , as well as at the quantum phase transition points in between . in the eh phase , at @xmath118 ,
most of the spectral weight is concentrated in the momentum range @xmath154 [ see fig .
[ skw](a ) ] .
the excitation gap appears at @xmath155 .
the dispersion of the maximum in @xmath6 behaves cosine - like for small - to - intermediate momenta , and is flattened close to the brillouin zone boundary ( above @xmath156 ) . with increasing @xmath108
the eh - oh transition occurs at @xmath157 , where the excitation gap closes at the momentum @xmath158 , as shown in fig .
[ skw](b ) .
deep in the haldane phase , the situation changes drastically [ see fig .
[ skw](c ) for @xmath159 .
now the dispersion of the maximum in @xmath6 takes a sine - like form .
again there are finite excitation gaps at @xmath158 ( haldane gap ) and @xmath160 .
this resembles the behavior found at the isotropic heisenberg point.@xcite here the spectral weight exclusively concentrates at @xmath161 and finite but small @xmath162 .
we finally ask whether the gap in @xmath6 again closes at the oh - afm transition if @xmath108 is increased further .
figure [ skw](d ) shows that the gap indeed closes , at @xmath163 , but this time at momentum @xmath164 , reflecting the lattice - period doubling in the afm phase .
obviously , @xmath6 follows the behavior of the neutral gap @xmath165 shown in fig .
[ gaps ] . in the afm phase
[ see fig .
[ skw](e ) with @xmath166 , the dispersion becomes flattened with a large excitation gap that opens at @xmath164 , however .
that is , the dynamical spin structure factor shows a distinct behavior in each phase of the spin-1 model with single - ion anisotropy .
interestingly , the results obtained in the eh , oh , and afm phases are similar to those for the mi , hi , and dw phases of the 1d ebhm.@xcite this corroborates that the spin-1 model can be taken as an effective model for the ebhm with @xmath152 .
we studied the topological properties of the anisotropic spin-1 @xmath0 model with single - ion anisotropy in close analogy to a recent investigation of the extended bose - hubbard model ( ebhm ) with a nearest - neighbor repulsion.@xcite the focus was on the nontrivial haldane phase as well .
the phase boundaries between trivial phases [ even haldane ( eh ) and afm phases ] and nontrivial odd haldane ( oh ) phase were determined numerically with high precision via the central charge .
the ground - state phase diagram resembles those of the restricted ebhm with a maximum number of bosons per site @xmath152 , but the topological phase takes a much narrower region in the parameter space . simulating the spin and neutral gaps , which correspond to the single - particle respectively neutral gaps in the ebhm
, we confirmed the closing of the gap at the trivial - nontrivial quantum phase transition as for the ebhm .
the degeneracy of the lowest entanglement level in the oh phase could be observed by finite - system dmrg calculations with both periodic ( p ) and open ( o ) boundary conditions ( bc ) .
with pbc the lowest level in the entanglement spectrum is fourfold degenerate in the oh phase ; notably , the system - size dependence of the results is much stronger than for obc . adopting half spins ( @xmath36 ) at the open edges ,
the twofold degeneracy corresponding to a single artificial edge in the entanglement calculations can be detected easily .
this degeneracy will be lifted turning on a finite perturbation that breaks the inversion symmetry of the lattice , independently from the bc used .
we furthermore used the dynamical dmrg technique to examine the dynamical spin structure @xmath6 which mimics the dynamical density fluctuations in the ebhm . in the topological @xmath167
oh phase a sinus - shaped dispersion was observed for finite anisotropy @xmath10 just as for the isotropic heisenberg model and the haldane insulator state of the ebhm .
moreover , @xmath6 shows a significant different momentum and energy dependence in three different gapful phases for both the spin-1 model and the ebhm .
we finally note that the influence of the particle - hole symmetry - breaking term @xmath33 in eq . on the properties of the constrained ebhm is almost negligible , not only for static but also for dynamical quantities .
the authors would like to thank y. fuji , f. ghmann , f. lange , s. nishimoto , and f. pollmann for valuable discussions .
this work was supported by deutsche forschungsgemeinschaft ( germany ) , sfb 652 , b5 .
.first excitation gap @xmath168 in the spin-1 @xmath0 chain as obtained by qmc , ed , and dmrg for a system size @xmath39 , at temperature @xmath169 , using the specified boundary conditions bc . [ cols="^,^,^,^,^",options="header " , ] after haldane s conjecture about the finite excitation gaps for integer - spin chains,@xcite it was a challenging issue to estimate these so - called haldane gaps numerically ( note that even the spin-1 @xmath0 heisenberg chain is not integrable ) .
white presented the first accurate dmrg results for the haldane gap,@xcite and subsequently a series of more elaborated dmrg,@xcite qmc @xcite and ed @xcite studies have been performed .
however , only obc have been used within the dmrg framework so far , mainly because of the smaller computational costs . in this appendix
, we demonstrate at least for the spin-1 heisenberg model that the haldane gap can also be determined using pbc , and the system - size dependence of the gap is much smaller than those with obc adopting the half spin edges [ cf .
[ haldane - state](b ) ] .
hence any finite - size scaling is needless .
figure [ fig : haldane - gap ] presents the finite - size extrapolation of the corresponding spin and neutral excitation gaps , as defined in sec .
[ sec : gap ] , for both obc and pbc .
the spin and neutral gaps become equal ( @xmath168 ) only at the heisenberg point for @xmath71 [ cf .
[ gaps](a ) ] .
computing @xmath168 for systems with up to @xmath170 sites and obc , we can extrapolate the results to the thermodynamic limit and obtain @xmath171 ( in agreement with ref . [ ] ) .
on the other hand , the first excitation gaps @xmath172 obtained with pbc and up to @xmath173 density - matrix states show almost no finite - size dependence ; see also the raw data for @xmath174 and @xmath90 in table [ tab : haldane - gap ] .
this value is very close to the ( low - temperature ) qmc@xcite and ed@xcite results and shows a perfect agreement with the very recent non - abelian dmrg data with obc.@xcite let us emphasize that although the accessible system size is rather limited for pbc , @xmath168 for pbc is always lower than for obc .
most notably , the system - size dependence is almost negligible ( for enough large @xmath39 ) , so that sophisticated extrapolation techniques or the use of special boundary conditions,@xcite are no longer mandatory for analyzing the haldane gap in spin-1 chains .
40ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty link:\doibase 10.1103/physrevlett.50.1153 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.56.371 [ * * , ( ) ] link:\doibase 10.1209/0295 - 5075/3/8/013 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.69.3571 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.59.799 [ * * , ( ) ] link:\doibase 10.1103/physrevb.45.304 [ * * , ( ) ] link:\doibase 10.1103/physrevb.67.104401 [ * * , ( ) ] link:\doibase 10.1088/0305 - 4470/28/19/003 [ * * , ( ) ] link:\doibase 10.1103/physrevb.80.155131 [ * * , ( ) ] link:\doibase 10.1103/physrevb.81.064439 [ * * , ( ) ] link:\doibase 10.1143/jpsj.80.043001 [ * * , ( ) ] link:\doibase 10.1088/1742 - 6596/302/1/012014 [ * * , ( ) ] link:\doibase 10.1088/1742 - 6596/320/1/012018 [ * * , ( ) ] link:\doibase 10.1103/physrevb.86.024403 [ * * , ( ) ] link:\doibase 10.1103/physrevb.85.075125 [ * * , ( ) ] link:\doibase 10.1103/physrevb.87.235106 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.97.260401 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.113.020401 [ * * , ( ) ] link:\doibase 10.1103/physrevb.77.245119 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1088/0953 - 4075/46/8/085303 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevb.34.6372 [ * * , ( ) ] link:\doibase 10.1103/physrevb.40.4709 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.101.010504 [ * * , ( ) ] @noop * * , @noop * * , ( ) link:\doibase 10.1103/physrevb.50.3045 [ * * , ( ) ] link:\doibase 10.1103/physrevb.77.134437 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.76.4038 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/j.physb.2014.03.007 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.224402 [ * * , ( ) ] link:\doibase 10.1143/jpsj.66.3379 [ * * , ( ) ] link:\doibase 10.1143/jpsj.66.3944 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.87.047203 [ * * , ( ) ] link:\doibase 10.1103/physrevb.48.3844 [ * * , ( ) ] link:\doibase 10.1143/jpsj.78.014003 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.054446 [ * * , ( ) ] | we reexamine the one - dimensional spin-1 @xmath0 model with on - site uniaxial single - ion anisotropy as to the appearance and characterization of the symmetry - protected topological haldane phase . by means of large - scale density - matrix renormalization group ( dmrg ) calculations
the central charge can be determined numerically via the von neumann entropy , from which the ground - sate phase diagram of the model can be derived with high precision .
the nontrivial gapped haldane phase shows up in between the trivial gapped even haldane and nel phases , appearing at large single - ion and spin exchange interaction anisotropies , respectively .
we furthermore carve out a characteristic degeneracy of the lowest entanglement level in the topological haldane phase , which is determined using a conventional finite - system dmrg technique with both periodic and open boundary conditions . defining the spin and neutral gaps in analogy to the single - particle and neutral gaps in the
intimately connected extended bose - hubbard model , we show that the excitation gaps in the spin model qualitatively behave just as for the bosonic system .
we finally compute the dynamical spin structure factor in the three different gapped phases and find significant differences in the intensity maximum which might be used to distinguish these phases experimentally . |
in the recent years the theory of consecutive pattern avoidance for permutations has experienced a rapid development since the publication of the important paper @xcite . among the latest publications one should mention e.g. @xcite , @xcite , @xcite , @xcite where a number of special cases was treated and the corresponding exponential generating function was explicitly found .
the present text is devoted to the same topic and is an extension of the application of homological methods to this theory initiated in @xcite .
we investigate a natural analog of the notion of wilf equivalence for consecutive pattern avoidance and obtain a rather general sufficient condition guaranteeing that this natural analog of wilf equivalence holds .
most of the definitions below are borrowed from @xcite and are ( hopefully ) standard in this area .
a _ permutation of length @xmath0 _ is a sequence containing each of the numbers @xmath1 exactly once . to every sequence @xmath2 consisting of @xmath0 distinct positive integers , we associate its _ standardization @xmath3 $ ] _ which is the permutation of length @xmath0 uniquely determined by the condition that @xmath4 if and only if @xmath3(i)<\operatorname{st}[s](j)$ ] . in other words ,
@xmath3 $ ] is the unique permutation of length @xmath0 whose relative order of entries is the same as that of @xmath2 .
for example , @xmath5=(231).$ ] in what follows we will refer to separate integers forming a permutation as its _
entries_. we say that a permutation @xmath6 of length @xmath0 _ contains a permutation @xmath7 of length @xmath8 as a consecutive pattern _ if for some @xmath9 the standardization @xmath10 $ ] coincide with @xmath7 .
if @xmath6 contains @xmath7 as a consecutive pattern we say that _ @xmath7 divides @xmath6 _ and use the notation @xmath11 . if @xmath11 and @xmath12 ( resp
@xmath13 ) we say that _
@xmath7 is a left ( resp .
right ) divisor of @xmath6_. the main notion in the theory of pattern avoidance for permutations is as follows .
we say that _ a permutation @xmath6 avoids a given permutation @xmath7 as a consecutive pattern _ if @xmath6 is not divisible by @xmath7 .
( throughout this paper we only deal with consecutive patterns , so the word `` consecutive '' will be omitted . )
the central problem of the theory of pattern avoidance is to count the number of permutations of a given length avoiding a given collection @xmath14 of forbidden patterns or , more generally , containing a given number of occurrences of patterns from @xmath14 .
this problem naturally leads to the following equivalence relation on collections of patterns defined in the simplest case by h. wilf , @xcite .
two collections of patterns @xmath15 and @xmath16 are said to be _ wilf equivalent _ if for every positive integer @xmath0 , the number of @xmath15-avoiding permutations of length @xmath0 is equal to the number of @xmath16-avoiding permutations of length @xmath0 .
the next notion although very natural seems to be new .
we say that two collections of patterns @xmath15 and @xmath16 are _ strongly wilf equivalent _ if for every positive integer @xmath0 and every nonnegative integer @xmath17 , the number of permutations of length @xmath0 with @xmath18 occurrences of patterns from @xmath15 equals the number of permutations of length @xmath0 with @xmath18 occurrences of patterns from @xmath16
. however , standard wilf equivalence deals with all pattern occurrences ( and not just consecutive ones ) . in the set - up of consecutive pattern avoidance we will speak about _
c - wilf equivalent_ ( resp . _
strongly c - wilf equivalent _ ) collections where `` c '' stands for consecutive .
( we use the notation : @xmath19 for strongly c - wilf equivalent collections . ) throughout this paper we assume that every collection of patterns @xmath14 is _ reduced _ , i.e. no two permutations @xmath20 are divisible by one another .
( notice that if @xmath21 then @xmath22 is strongly c - wilf equivalent to @xmath14 . ) following @xcite consider two natural exponential generating functions in one and two variables respectively : @xmath23 associated to a given collection of patterns @xmath14 . here
@xmath24 ( resp .
@xmath25 ) is the number of permutations of length @xmath0 avoiding all patterns from @xmath14 ( resp .
the number of permutations of length @xmath0 with exactly @xmath18 occurrences of patterns from @xmath14 ) . obviously , @xmath26 .
( recall that we only count consecutive occurrences . )
hilbert series very similar to @xmath27 and @xmath28 are often considered in the theory of associative algebras .
the well - known method of their study is based on the so - called bar - cobar duality which roughly means that a graded associative algebra @xmath29 and the @xmath30-coalgebra @xmath31 are dual with respect to the functor @xmath32 . as a corollary of this duality
one gets that the hilbert series of @xmath29 and of @xmath31 are the inverses of each other , i.e. their product equals @xmath33 .
( see @xcite for the details on different computational methods for the hilbert series of associative algebras and their homology . )
it seems highly plausible that for an associative algebra with few relations a combinatorial description of its homology is simpler than that of the algebra itself .
however , for algebras with many relations the situation is the opposite one .
recall that the set of permutations avoiding an arbitrary fixed collection @xmath14 has an important additional structure , see appendix in @xcite .
namely , in monoidal category it can be considered as the monomial basis of an algebra with monomial relations .
( we refer an interested reader to the above appendix in @xcite and references therein for the details . in particular , one can find the definition of the homology functor in the latter appendix . )
therefore , it seems natural to use the above mentioned homological duality in the theory of pattern avoidance .
combinatorial data appearing in this context is based on a generalization of the so - called cluster method of i. goulden and d. jackson , @xcite .
we explain below how one can get combinatorial information ( for example , about the coefficients of the generating functions ) of the corresponding graded homological vectorspaces for collections of patterns with few entries . to describe our results we need to recall the definition of a combinatorial gadget called _ clusters _ in @xcite .
they generalize the notion of a linkage given below .
a permutation @xmath6 of length @xmath0 is called a _ linkage _ of an ordered pair of ( not necessarily distinct ) patterns @xmath34 of lengths @xmath35 and @xmath36 if ( i ) @xmath37 ; and ( ii ) the standardizations @xmath38 $ ] and @xmath39 $ ] are equal to @xmath7 and @xmath40 resp .
since the length of @xmath6 is less than the sum of the lengths of @xmath7 and @xmath40 one has that the standardizations of the right truncation of @xmath7 and left truncations of @xmath40 should be the same . setting @xmath41
we say that a pair @xmath42 has a _ @xmath43-overlapping _ or _
@xmath43-overlaps_. in other words , a pair @xmath34 @xmath43-overlaps whenever the standardization @xmath44 $ ] is equal to the standardization @xmath45 $ ] .
( notice that there might be several different linkages of two given patterns @xmath7 and @xmath40 of the same length . )
a cluster is a way to link together several patterns from a given set .
more precisely , a @xmath18-cluster w.r.t .
a given collection of patterns @xmath14 is a triple @xmath46 where @xmath6 is a permutation , @xmath47 is a list of ( not necessarily distinct ) patterns from @xmath14 , and @xmath48 is a list of positive integers such that * for every @xmath49 , @xmath50=\pi_j\in \varpi$ ] , where @xmath51 is the length of @xmath52 ( here @xmath53 labels the beginning of the pattern @xmath52 in @xmath6 ) ; * @xmath54 ( patterns are listed from left to right ) and @xmath55 ( adjacent patterns are linked ) ; * @xmath56 , and the length of @xmath6 is equal to @xmath57 ( i.e. @xmath6 is completely covered by the patterns @xmath58 ) .
denote by @xmath59 the number of @xmath18-clusters of length @xmath0 in a collection @xmath14 and introduce the exponential generating function @xmath60 ( here we use a natural convention that there always exists exactly one ( fictitious ) @xmath61-cluster and , therefore , the above generating function starts with @xmath62 . )
the next result is an immediate consequence of the general cluster method of i. goulden and d. jackson , @xcite and its homological proof for the case of permutations can be found in @xcite .
[ th : gj ] in the above notation one has : @xmath63 the exponential generating function @xmath64 of the numbers of permutations avoiding patterns from a given collection @xmath14 satisfies the relation : @xmath65 in general , the problem of counting the number of @xmath18-clusters in a given collection of patterns @xmath14 does not seem to be easier than counting the number of permutations of a given length avoiding @xmath14 .
on the other hand , there exist natural classes of collections for which counting @xmath18-clusters is an easier task , see 3 .
one can guess that since clusters can be described in terms of linkages of pairs of patterns the number of clusters can also be determined in terms of the combinatorics of these linkages .
exploiting the latter idea we were able to prove the following ( main ) result of this note .
[ th : main ] two collection of patterns @xmath15 and @xmath16 are strongly c - wilf equivalent if there exists a bijection @xmath66 preserving the following three properties : * ( lengths ) : for any @xmath67 its length equals to that of @xmath68 ; * ( linkages ) : a pair of patterns @xmath34 from @xmath15 has a linkage of length @xmath0 if and only if the pair of its images @xmath69 from @xmath16 has a linkage of the same length @xmath0 . *
( overlapping sets ) : for each overlapping of any pair of patterns from @xmath15 the bijection @xmath70 preserves the subsets of entries that overlaps .
more precisely , for any pair @xmath42 of patterns @xmath71 of lengths @xmath35 and @xmath36 resp . and
an arbitrary positive integer @xmath72 the coincidence of the standardizations @xmath73 = \operatorname{st}[(\pi'(1)\ldots\pi'(k))]$ ] implies the coincidence of sets : @xmath74 [ theorem_wilf_overaping ] the simplest case where theorem [ th : main ] applies is to collections with a single pattern having no nontrivial self - overlappings . the following result implied by theorem [ th : main ]
was first conjectured by s. elisalde in @xcite and later proved in @xcite by homological methods and , simultaneously , by j. remmel whose methods were based on @xcite .
namely , two collections of patterns each containing a single permutation without nontrivial self - overlappings are strongly c - wilf equivalent if * the lengths of the permutations coincide ; * the first entry and resp . the last entry of the permutations coincide .
a series of particular examples covered by theorem [ th : main ] can be found in 5 of @xcite .
namely , the following definitions are borrowed from the latter paper .
we say that a pair of permutations @xmath75 and @xmath76 _ has a separation property _
if @xmath77 avoids the permutation @xmath78 and @xmath79 avoids @xmath80 . with each pair of permutations @xmath75 , @xmath81 and a natural number @xmath35
one can associate the subset @xmath82 of permutations defined by two following properties .
we say that @xmath83 iff * the standardizations of the @xmath43 first and @xmath84 last entries coincide with @xmath79 , and @xmath77 resp . *
the @xmath43 first entries are strictly smaller than the @xmath84 last entries ; the @xmath84 last entries are strictly smaller than the remaining entries of @xmath7 in the middle . in other words , @xmath85 for any triple of indices @xmath86 such that @xmath87 .
fix a pair of permutations @xmath79 and @xmath77 having a separation property and a @xmath88-tuple of natural numbers @xmath89 .
then all collections of @xmath88 distinct patterns @xmath90 such that @xmath91 are strongly wilf equivalent . for all @xmath92 the @xmath93 entries in the middle of each pattern
@xmath94 never appear in the overlapping sets .
let us present a few more examples illustrating how our theorem works in practice .
the following patterns @xmath95 are pairwise c - wilf equivalent .
they have self - overlappings of lengths @xmath33 and @xmath96 and coinciding pairs of the first two and the last two entries .
the next pair of c - wilf equivalent patterns @xmath97 have self - overlappings of lengths @xmath33 and @xmath98 and the corresponding subsets of their initial and final entries of lengths 1 and 4 coinciding while their initial and final subwords are different . finally , here is an example @xmath99 of c - wilf equivalent collections with 2 patterns in each . in
2 we prove theorem [ th : main ] and in 3 we apply our main construction to a class of collections of patterns and obtain a system of linear ordinary differential equations satisfied by @xmath100 together with a set of similar generating functions defined below . in the follow - up @xcite of the present paper we plan to study different asymptotic properties of @xmath101 using the suggested approach .
_ acknowledgements .
_ the authors are sincerely grateful to s. kitaev for the e - mail correspondence concerning this subject .
research of the first author is supported by rfbr 10 - 01 - 00836 , rfbr - cnrs-10 - 01 - 93111 , rfbr - cnrs-10 - 01 - 93113 , and by the federal programm under the contract 14.740.11.0081 .
our proof of theorem [ theorem_wilf_overaping ] consists of an algorithm computing the cluster generating function @xmath102 of a given collection of patterns @xmath14 .
it will then be relatively easy to see that this algorithm uses only the lengths and the overlapping subwords for pairs of patterns from @xmath14 considered as sets . to start with ,
we define for an arbitrary collection of patterns @xmath14 a certain directed graph with labelled vertices and edges .
the important circumstance is that the number of @xmath18-clusters with fixed initial and final subwords will be equal to the number of properly weighted paths of length @xmath18 in this graph with fixed initial and final vertices .
the required weights can be computed using the edge labels . as a consequence
we get that this graph uniquely defines the generating functions @xmath102 and , therefore , @xmath28 , see theorem [ th : gj ] .
namely , given an arbitrary collection of patterns @xmath14 define its directed graph @xmath103 with labelled vertices and edges as follows .
the vertices of @xmath103 will be labelled by permutations ( of , in general , different lengths ) and the labels of the edges are defined below . * to define
the vertices assume that some permutation @xmath104 is a left divisor of a pattern @xmath105 and , at the same time , a right divisor of a ( not necessarily different ) pattern @xmath106 .
then we assign to @xmath104 a vertex @xmath107 of @xmath103 and , naturally , label this vertex by @xmath108 .
( notice that the same @xmath108 can arise from different pairs @xmath109 .
in particular , the trivial @xmath33-element permutation @xmath110 comes from an arbitrary pair of not necessarily distinct patterns .
@xmath111 is called the _
distinguished vertex _ of @xmath103 and the set of all vertices of @xmath103 is denoted by @xmath112 . )
* to define the edges take a pattern @xmath113 of some length @xmath35 and a pair @xmath114 of its initial and final subwords of lengths @xmath43 and @xmath84 ( i.e. @xmath115 and @xmath116 ) such that standardizations @xmath117,\operatorname{st}[\pi_f]$ ] are the vertices of @xmath103 .
let @xmath118 and @xmath119 be the subsets of entries that appears in @xmath120 and @xmath121 resp .
( i.e. @xmath122 and @xmath123 ) .
the triple @xmath124 then defines the directed edge from the vertex @xmath117 $ ] to the vertex @xmath125 $ ] which we label by the triple @xmath126 .
notice that @xmath118 and @xmath119 are considered as _
unordered sets_. _ notation .
_ in the present text we show all vertices of @xmath103 as encircled permutations to distinguish them from their labeling permutations
. we will try to keep our notation straight throughout the whole text by denoting similar quantities by the same letter and adding extra indices if required .
for example , @xmath35 will typically mean the length of a pattern @xmath7 from a collection , @xmath43 should denote the length of a permutation @xmath108 which labels a vertex of @xmath103 originating from a @xmath43-overlapping , @xmath0 will stand for the length of a cluster .
four examples of @xmath103 are given below .
the upper left example is constructed from the collection @xmath127 of two patterns with no nontrivial overlappings .
the upper right example comes from the single pattern @xmath128 having self - overlappings of lengths @xmath33 and @xmath129 .
the meaning of two other examples should be clear now . .
] our main technical result is as follows .
[ th : graph ] the graph @xmath103 uniquely determines the generating function @xmath102 .
the next corollary immediately implies theorem [ theorem_wilf_overaping ] . two collections of patterns @xmath15 and @xmath16 having isomorphic graphs @xmath130 and @xmath131 are strongly c - wilf - equivalent .
( here by an isomorphism we mean a graph isomorphism preserving the labels of edges . the labels of vertices can change . ) [ algorithm ] to prove theorem [ th : graph ] we present a natural algorithm calculating the number of @xmath18-clusters in a given collection @xmath14 in terms of its graph @xmath103 .
namely , each vertex @xmath107 and a positive integer @xmath0 uniquely determine the subset @xmath132 consisting of all @xmath18-clusters @xmath46 , such that the length of @xmath6 is equal to @xmath0 and the standardization of the initial subword of @xmath6 is equal to @xmath108 .
moreover , with each word @xmath133 of length @xmath43 ( where @xmath43 is the length of @xmath108 ) one can associate the subset @xmath134\subset cl_{v , n , q}$ ] consisting of those clusters in @xmath132 which have @xmath135 as their initial subword .
we will explain how one should compute the cardinalities of @xmath134 $ ] by induction on @xmath18 using the edge labels in @xmath103 .
therefore , the cardinalities of @xmath132 can also be computed inductively as the sums over different @xmath135 . since the standardization of any word of length @xmath33 equals @xmath110 the set @xmath136 coincides with the set of all @xmath18-clusters of length @xmath0 .
( the cardinality of the latter set is one of required coefficients in the cluster generating function @xmath102 . )
let us now return to the induction step .
take an arbitrary vertex @xmath137 and let @xmath138
, @xmath139 be the list of all edges in @xmath103 starting at the vertex @xmath107 .
denote by @xmath140 the length of permutation @xmath141 labeling the vertex @xmath142 and denote by @xmath51 the length of the pattern @xmath52 .
we present below a recurrence relation expressing the cardinality @xmath143 $ ] of the set @xmath134 $ ] in terms of the cardinalities @xmath144 $ ] of @xmath145 $ ] with the summation taken over a certain subset of words @xmath146 . using this relation
we can calculate inductively each @xmath143 $ ] and then obtain the required @xmath147 by summation over different @xmath135 .
it will be convenient to subdivide the sets @xmath132 and @xmath134 $ ] into the subsets indexed by the edges starting at the vertex @xmath107 .
for example , @xmath148 is the subset of @xmath18-clusters formed by linkages of length @xmath0 between the pattern @xmath52 and a @xmath149-cluster from @xmath150 .
one has @xmath151=v \end{smallmatrix } } cl_{v , n , q}[p_1\ldots p_k ] = \sum _ { \begin{smallmatrix } { 1\leq p_1,\ldots , p_k \leq n , } \\
st[(p_1\ldots p_k)]=v \end{smallmatrix } } \sum_{j=1}^d cl_{v\stackrel{\pi_j}{\mapsto } v_j , n , q}[p_1\ldots p_k ] .
\label{add_statistics}\ ] ] therefore , it is sufficient to find recurrence relations expressing the terms @xmath152 $ ] in the right - hand side of using @xmath153 $ ] . to avoid very cumbersome notation
let us take the case of a single edge starting at @xmath107 which is equivalent to fixing @xmath141 in the above formulas .
let @xmath154 be an edge in a graph @xmath103 coming from a pattern @xmath7 of length @xmath35 and let @xmath43 and @xmath84 be the lengths of the permutations labeling @xmath107 and @xmath155 resp . to explain our recurrence we need to introduce the following extra notation associated to @xmath7 .
if @xmath156 let @xmath157 be the permutation which is the inverse of the standardization of the @xmath43 first and the @xmath84 last entries of @xmath7 , and let @xmath158 be the composition of @xmath159 with the shifting map @xmath160 defined by the formula : @xmath161 in other words , @xmath158 prescribes the rule how to write down the @xmath43 first and the @xmath84 last entries of the pattern @xmath7 in the increasing order : @xmath162 the next statement gives the required recurrence .
[ lm:1 ] the following relations hold : * for @xmath156 set @xmath163 $ ] . then @xmath164= \\ \sum _ { \begin{smallmatrix } p_{k+1},\ldots , p_{k+k'}\ : \\
st[(p_1\ldots p_{k+k'})]= \tilde{\pi } \end{smallmatrix } } \binom{p_{\psi(1)}-1}{\pi(\overline{\psi}(1))-1}\times \left [ \prod_{j=1}^{k+k'-1 } \binom{p_{\psi(j+1)}-p_{\psi(j)}-1}{\pi(\overline{\psi}(j+1))-\pi(\overline{\psi}(j))-1 } \right ] \times \binom{n - p_{\psi(k+k')}}{l-\pi(\overline{\psi}(k+k'))}\times \\ \times cl_{v',n - l+k',q-1}[p_{k+1}-\pi(l - k'+1 ) + v'(1),\ldots , p_{k+k'}-\pi(l)+v'(k ' ) ] .
\label{one_edge}\end{gathered}\ ] ] * for @xmath165 one has : @xmath166= } \\ { = \sum_{p_{k+1},\ldots , p_{l } : \operatorname{st}[(p_1\ldots p_l)]= \pi } cl_{v',n - l+k',q-1}[p_{l - k'+1}-\pi(l - k'+1 ) + v'(1),\ldots , p_{l}-\pi(l)+v'(k ' ) ] . }
\label{one_edge_1}\end{gathered}\ ] ] the range of summation in can be easily derived from our convention on the binomial coefficients claiming that @xmath167 if either @xmath168 or @xmath169 .
moreover , we assume that @xmath170 s are pairwise different positive integers not exceeding @xmath0 . for the induction base we use the following initial data : @xmath171 we show how to prove . in formula one has the summation over all patterns @xmath172 such that the word @xmath173 is fixed and coincides with @xmath174 .
indeed , the numbers @xmath170 are ordered by the permutation @xmath159 as follows : @xmath175 .
therefore , there are @xmath176 choices of entries less than @xmath177 among the first @xmath35 entries of @xmath6 ; there are @xmath178 choices of entries greater than @xmath177 and less than @xmath179 , @xmath180 ; there are @xmath181 choices of entries greater than @xmath182 among the first @xmath35 entries of @xmath6 ; and @xmath183 $ ] ways to choose the remaining standardization of the last @xmath184 entries of @xmath6 . in the union of the @xmath43 initial entries and the @xmath84 final entries of @xmath7
covers the whole list of entries of @xmath7 , i.e. the set @xmath185 .
therefore , all binomial coefficients appearing in should be equal to @xmath33 which leads to .
as an immediate consequence of lemma [ lm:1 ] one can see that the numbers @xmath186 $ ] of @xmath187-clusters depend only on the length , the @xmath43 first and the @xmath84 last entries of @xmath7 considered as sets .
this justifies the information we use as the edge labels of the graph @xmath103 .
the formulas expressing @xmath188 $ ] in terms of @xmath189 $ ] depend only on the labeling of the edges starting at @xmath107 .
therefore , these cardinalities can be computed by induction on @xmath18 using the edge labels of the graph @xmath103 .
finally , as we mentioned before , the set of all @xmath18-clusters of length @xmath0 of the whole collection @xmath14 is equal to the set @xmath136 .
let us consider separately the situation when @xmath14 contains just a single pattern since in this case some substantial simplifications of our construction can be done .
first of all the following observation explains why the graph @xmath190 is not required .
let @xmath7 be a pattern of length @xmath35 and let @xmath191 ,@xmath192 be the list of all distinct lengths of possible self - linkages of @xmath7 , i.e. @xmath193 ,
, @xmath194 is the list of distinct lengths of self - overlappings of @xmath7 .
then @xmath190 is a complete directed graph on @xmath88 vertices with loops and with lengths of the underlying permutations being equal to @xmath195 .
each ordered pair of ( not necessary distinct ) vertices of @xmath190 are connected by exactly one directed edge labeled by the corresponding initial and final subwords of @xmath7 .
[ one_pattern ] it is obvious that @xmath196 and denote by @xmath43 ( @xmath197 ) the length of the largest overlapping .
let @xmath198 be the standardization of the @xmath199 first entries of @xmath7 ( i.e. @xmath198 is the labeling permutation of the @xmath2-th vertex in @xmath190 ) .
since all patterns involved in any cluster coincide with @xmath7 the standardization of the initial subword of any cluster should always be the same .
hence for different @xmath198 and fixed @xmath0 and @xmath18 all the sets @xmath200 coincide .
therefore , it makes sense to denote by @xmath201 and @xmath202 $ ] the cardinalities of the set of @xmath18-clusters of length @xmath0 and those having @xmath203 as their initial subword resp .
we introduce the same set of notation for the self - overlappings of @xmath7 similar to what have been using in lemma [ lm:1 ] for the case @xmath204 .
namely , for @xmath205 let @xmath206 be the permutation which is the inverse of the standardization of the @xmath43 first and the @xmath199 last entries of @xmath7 ; for @xmath207 let @xmath159 be the inverse of @xmath7 .
let @xmath208 be the composition @xmath209 using which one gets the following rearrangement of the first @xmath43 and last @xmath199 elements of @xmath7 in the increasing order : @xmath210 additionally ,
let @xmath211 be the standardization of the @xmath43 first and @xmath199 last entries of @xmath7
. in the case of a single pattern lemma [ lm:1 ] implies the next result .
[ lm:2 ] for a single pattern the recurrence formula for the numbers of @xmath18-clusters is as follows : @xmath212 = \\ \sum _ { \begin{smallmatrix } s \ : \\
k_s < l - k \end{smallmatrix } } \sum _ { \begin{smallmatrix } 1\leq p_{k+1},\ldots , p_{2k}\leq n , \\
st[(p_1\ldots p_{k+k_s})]= \tilde{\pi}_s \end{smallmatrix } } \binom{p_{\psi_s(1)}-1}{\pi(\overline{\psi_s}(1))-1}\times \left [ \prod_{j=1}^{k+k_s-1 } \binom{p_{\psi_s(j+1)}-p_{\psi_s(j)}-1}{\pi(\overline{\psi_s}(j+1))-\pi(\overline{\psi_s}(j))-1 } \right ] \times \binom{n - p_{\psi_s(k+k_s)}}{l-\pi(\overline{\psi_s}({k+k_s}))}\times \\
\times cl_{n - l+k_s , q-1}[p_{k+1}-\pi(l - k_s+1 ) + v_s(1),\ldots , p_s({k+k_s})-\pi(l)+v_s(k_s),p_{k+k_s+1},\ldots , p_{2k } ] + \\ + \sum _ { \begin{smallmatrix } s \ :
\\ k_s \geq l - k \end{smallmatrix } } \sum _ { \begin{smallmatrix } 1\leq p_{k+1},\ldots , p_{l+k - k_s}\leq n , \\ st[(p_1\ldots p_{l})]= \pi \end{smallmatrix } } cl_{n - l+k_s , q-1}[p_{l - k_s+1}-\pi(l - k_s+1 ) + v_s(1),\ldots , p_{l}-\pi(l)+v_s(k_s),p_{l+1},\ldots , p_{l+k - k_s } ] .
\label{recurent - one - pattern } \end{gathered}\ ] ] ( as above we assume that the set of @xmath61-clusters contains the unique fictitious element of length @xmath33 while the set of @xmath33-clusters contains the single pattern @xmath7 . )
in this section we discuss a specific class of collection of patterns .
our method from
[ sec : proofs ] allows us to construct a system of linear ordinary differential equations in the variable @xmath62 for the cluster generating functions @xmath102 together with a set of similar generating functions associated to a given collection of patterns @xmath14 .
the main definition is as follows .
a collection of patterns @xmath14 is called _ monotone _ if for all @xmath213 and for each pair ( not necessarily distinct ) patterns @xmath42 from @xmath14 the existence of their @xmath43-overlapping implies that the initial subword of the pattern @xmath40 does not contain entries greater than @xmath43 .
[ lm : monot_link ] let @xmath6 be a linkage of a pair of patterns @xmath42 .
suppose that the initial subword of @xmath7 of length @xmath43 does not contain entries greater than @xmath43 and that the initial subword of @xmath40 of length @xmath84 does not contain entries greater than @xmath84 ( where @xmath84 is the length of the overlapping of the pair @xmath42 in @xmath6 ) .
then the initial subword of length @xmath43 of @xmath6 should be equal to the initial subword of @xmath7 , i.e. @xmath214 for @xmath215 .
let @xmath35 and @xmath36 be the lengths of @xmath7 and @xmath40 resp . then the length of @xmath6 should be equal to @xmath216 .
set @xmath217 and @xmath218 respectively .
the number @xmath219 is the minimal entry of @xmath7 and should be less than or equal to @xmath43 ( @xmath220 respectively ) .
therefore , @xmath221 should be the smallest entry among the @xmath35 first entries of the linkage @xmath6 and @xmath222 should be the smallest entry among the @xmath36 last entries of @xmath6 .
in particular , @xmath223 .
this implies that @xmath221 is the smallest entry in the whole @xmath6 and hence @xmath224 .
similar arguments based on simple properties of standardizations imply that @xmath225 is the only entry in @xmath6 greater than @xmath33 etc .
lemma [ lm : monot_link ] implies that for an arbitrary vertex @xmath226 all clusters in the subset @xmath132 should always have @xmath108 as their initial subword .
this means that the number @xmath227 $ ] is non - vanishing if and only if @xmath228 , where @xmath43 is the length of @xmath108 .
consider two vertices @xmath107 and @xmath229 of lengths @xmath43 and @xmath84 connected by an edge @xmath230 in @xmath103 .
we can simplify the formula using our assumption on the intersections of clusters .
at first consider the case when the length @xmath35 of the pattern @xmath7 is greater than @xmath231 .
one observes that in this case there is no summation in since the summands in the left - hand and the right - hand sides of are non - vanishing if and only if for @xmath215 one has that @xmath232 and for @xmath233 one has that @xmath234 therefore , only the last binomial coefficient in the product in the right - hand side of is different from @xmath33 .
denote by @xmath235 the maximal entry of the final subword of @xmath7 .
( in notation of lemma [ lm:1 ] we have @xmath236 . )
since we do not have to specify the arguments of the functions @xmath188 $ ] the resulting recurrence relation for the cardinalities of the set of clusters is : @xmath237 the case @xmath165 is also covered by since in this case the maximal entry @xmath235 should be equal to the length @xmath35 of the permutation @xmath7 and the corresponding binomial coefficient is equal to @xmath33 .
introduce the next family of generating functions : @xmath238 one for each vertex @xmath107 of @xmath103 .
let @xmath138 , @xmath139 be the list of all edges in @xmath239 starting at @xmath107 . denote by @xmath51 the length of the pattern @xmath52 ; denote by @xmath140 the length of @xmath141 , and , finally , denote by @xmath240 the maximal entry among the @xmath140 last entries of @xmath52 .
formula implies the following reccurence relation : @xmath241 [ th : monotone_patterns ] given a monotone collection of patterns @xmath14 one has that the cluster generating function @xmath242 together with all @xmath243 solve the following system of linear ordinary differential equations in @xmath62 : @xmath244 here @xmath245 and @xmath107 runs over the set @xmath239 of all vertices of @xmath103 .
( boundary conditions for each @xmath246 can be easily determined in each particular case using the initial terms in . ) as an immediate consequence of theorem [ th : monotone_patterns ] one gets that @xmath242 satisfies a certain high order linear ordinary differential equation with polynomial coefficients ( which can be obtained from the above system after the elimination of all @xmath247 . )
two monotone collections of patterns @xmath15 and @xmath16 are strongly c - wilf equivalent if there exists a bijection @xmath248 preserving the first two properties as in theorem [ th : main ] ( i.e. preserving lengths and linkages ) and , additionally , preserving the maxima of the overlapping sets . more precisely , for any pair @xmath42 of patterns @xmath71 of lengths @xmath35 and @xmath36 resp . and an arbitrary positive integer @xmath72 the coincidence of the standardizations @xmath73 = \operatorname{st}[(\pi'(1)\ldots\pi'(k))]$ ] implies the coincidence of the maxima of the sets : @xmath249 ( two other sets @xmath250 and @xmath251 should coincide with the standard set @xmath252 because of the monotonicity property . )
let us present further simplifications for monotone collections with a single pattern . as it was shown in [ section - one - pattern ]
we do not really need the graph @xmath103 for collections with one pattern . take a monotone pattern @xmath7 of length @xmath0 and let @xmath254 be the list of all different lengths of self - overlappings of @xmath7 .
let @xmath255 be the list of maxima among the corresponding number of the last entries of @xmath7 , i.e. @xmath256 .
given a monotone pattern @xmath7 one has the following recurrence formula for the set of its @xmath18-clusters of length @xmath0 : @xmath257 and its cluster generating function @xmath258 solves the linear ordinary differential equation : @xmath259 with the boundary conditions : @xmath260 .
[ lm : one - pat - dif - eq ] finally , using these considerations we can completely describe which permutations of length @xmath263 ( @xmath264 ) are c - wilf equivalent .
our main theorem [ th : main ] gives necessary and sufficient condition for c - wilf equivalence of two patterns of length @xmath263 .
notice that there are two natural transformations of patterns preserving the cluster generating functions .
the first one is the reversion that rewrites a pattern backwards , i.e. it sends ( @xmath265 to @xmath266 ) .
the second one takes the complement of a pattern , i.e. it sends @xmath267 to @xmath268 .
they generate the group @xmath269 acting on @xmath270 .
one can easily check that this action has @xmath271 orbits of which @xmath98 orbits with representatives @xmath272 have length @xmath96 and the remaining @xmath273 have length @xmath98 .
additionally , there are 14 orbits whose permutations have no nontrivial selp - overlappings ; 15 orbits with the only nontrivial self - overlapping of length @xmath96 ; 2 orbits with the only nontrivial self - overlapping of length @xmath129 , and a single orbit with self - overlappings of length 2 , 3 and 4 , see the lists of representatives in proposition [ pr : s5 ] . [
pr : s5 ] subdividing the representatives of the orbits of the latter group action on @xmath270 into the subsets according to the lengths of their maximal possible self - overlappings one gets the following : * c - wilf equivalent orbits having no nontrivial overlappings are enumerated by their first and last elements : @xmath274 @xmath275 * among all @xmath276 orbits with only 2-overlappings having the representatives @xmath277 no two are strongly c - wilf equivalent to each other ; * @xmath96 orbits with representatives @xmath278 having @xmath129-overlappings are not c - wilf equivalent ; * the unique orbit with overlappings of lengths @xmath279 is represented by the monotone pattern @xmath280 . besides the application of theorem [ th : main ] one has to check by hand that the number of @xmath129-clusters is distinguishing all the 15 orbits with only 2-overlappings and already the number of @xmath96-clusters distinguishes between the two orbits having only @xmath129-overlappings .
let us finish the paper by presenting the ( system of ) linear differential equations for our examples in fig.[graph - example ] .
one can easily check that all the collections @xmath281 shown there are monotone .
set @xmath282 and @xmath283 for @xmath15 and notice that its graph contains a single vertex .
then in the above notation one has @xmath284 .
equation for the cluster generating function @xmath285 then takes the form : @xmath286 for the collection @xmath287 that has self - overlappings of lengths @xmath33 and @xmath129 one has @xmath288 . with this data equation
takes the form : @xmath289 for the collection @xmath290 denote its edges by @xmath291 , @xmath292 , @xmath293 and @xmath294 .
then one has @xmath295 ; @xmath296 ; @xmath297 ; @xmath298 .
thus , one gets the following system of equations : @xmath299 in this case it is easy to get a linear ordinary differential equation satisfied by @xmath300 by dividing both sides of the first equation by @xmath301 , differentiating the result with respect to @xmath62 and equating the expressions for @xmath302 from the first and second equations .
the resulting equation has the form @xmath303 finally , for the collection @xmath304 denote its edges by @xmath305 ; @xmath306 ; @xmath307 ; @xmath308 ; @xmath309 , and @xmath310 .
then one has @xmath311 ; @xmath312 ; @xmath313 ; @xmath314 ; @xmath315 ; @xmath316 .
thus , one gets the following system of equations : @xmath317 let us extract from this system a linear ordinary differential equation satisfied by @xmath300 . to simplify our notation set @xmath318
then we get the system : @xmath319 since the coefficients in the left - hand sides are both equal to @xmath33 we can eliminate @xmath108 by differentiating the first equation a number of times and substituting @xmath320 from the second equation till we can solve both equations for the remaining @xmath108 and its derivative of appropriate order . in our concrete case , differentiating the first equation twice and substituting @xmath320 from the second equation we get @xmath321 differentiating the first equation in the latter system once and substituting @xmath320 again we get @xmath322 finally , equating the expressions for @xmath323 from the first equation in the latter system and from the original equation for @xmath324 we get the following equation @xmath325 containing @xmath326 and its derivatives only . substituting @xmath327
we obtain the required equation @xmath328 for the cluster generating function .
notice that since the leading terms in the left - hand sides of the system are always equal to @xmath33 the elimination process similar to the one just described will always lead to an equation satisfied by the cluster generating function @xmath28 . on the other hand
, there is no guarantee that the obtained linear ordinary equation with polynomial coefficients will have the minimal possible order among such equations satisfied by @xmath28 .
s. kitaev , a survey on partially ordered patterns . in permutation patterns ( 2010 ) , s. linton , n. ruskuc , and v. vatter , eds . ,
376 of london mathematical society lecture note series , cambridge university press , pp .
115 - 135 . | using the approach suggested in @xcite we present below a sufficient condition guaranteeing that two collections of patterns of permutations have the same exponential generating functions for the number of permutations avoiding elements of these collections as consecutive patterns . in short ,
the coincidence of the latter generating functions is guaranteed by a length - preserving bijection of patterns in these collections which is identical on the overlappings of pairs of patterns where the overlappings are considered as unordered sets .
our proof is based on a direct algorithm for the computation of the inverse generating functions . as an application we present a large class of patterns where this algorithm is fast and ,
in particular , allows to obtain a linear ordinary differential equation with polynomial coefficients satisfied by the inverse generating function . |
in the early 1900s , bernstein @xcite asked the following question : given a collection of subsets of a set @xmath3 , is there a partition of @xmath3 into @xmath4 such that no subset is contained in either @xmath5 or @xmath6 ?
if we think of the elements of @xmath3 as vertices and of each subset as a hyperedge , the question can be rephrased as whether a given hypergraph can be 2-colored so that no hyperedge is monochromatic . of particular interest
is the setting where all hyperedges contain @xmath0 vertices , @xmath0-uniform hypergraphs .
this question was popularized by erds who dubbed it `` property b '' in honor of bernstein and has motivated some of the deepest advances in probabilistic combinatorics . indeed , determining the smallest number of hyperedges in a non-2-colorable @xmath0-uniform hypergraph remains one of the most important problems in extremal graph theory , perhaps second only to the ramsey problem @xcite .
a more modern problem , with a somewhat similar flavor , is boolean satisfiability : given a cnf formula @xmath7 , is it possible to assign truth values to the variables of @xmath7 so that it evaluates to true ?
satisfiability has been the central problem of computational complexity since 1971 when cook @xcite proved that it is complete for the class np .
the case where all clauses have the same size @xmath0 is known as @xmath0-sat and is np - complete for all @xmath8 . for both @xmath0-sat and property b
it is common to generate random instances by selecting a corresponding structure at random . indeed ,
random formulas and random hypergraphs have been studied extensively in probabilistic combinatorics in the last three decades .
while there are a number of slightly different models for generating such structures uniformly at random " , we will see that results transfer readily between them . for the sake of concreteness , let @xmath9 denote a formula chosen uniformly among all @xmath10 formulas on @xmath11 variables with @xmath12 clauses .
similarly , let @xmath13 denote a hypergraph chosen uniformly among all @xmath14 hypergraphs with @xmath11 vertices and @xmath12 hyperedges .
we will say that a sequence of events @xmath15 occurs _ with high probability _
( ) if @xmath16 = 1 $ ] and _ with uniformly positive probability _
( ) if @xmath17 > 0 $ ] . throughout the paper
, @xmath0 will be arbitrarily large but fixed . in recent years , both problems have been understood to undergo a `` phase transition '' as the ratio of constraints to variables passes through a critical threshold .
that is , for a given number of vertices ( variables ) , the probability that a random instance has a solution drops rapidly from 1 to 0 around a critical number of hyperedges ( clauses ) .
this sharp threshold phenomenon was discovered in the early 1990s , when several researchers @xcite performed computational experiments on @xmath18 and found that while for @xmath19 almost all formulas are satisfiable , for @xmath20 almost all are unsatisfiable .
moreover , as @xmath11 increases , this transition narrows around @xmath21 . along with similar results for other fixed @xmath8
this has led to the following popular conjecture : * satisfiability threshold conjecture : * _ for each @xmath22 , there exists a constant @xmath23 such that _ @xmath24= \begin{cases } 1 & \mbox{if $ r < r_k$}\\ 0 & \mbox{if $ r > r_k$ } \enspace . \end{cases}\ ] ] in the last ten years , this conjecture has become an active area of interdisciplinary research , receiving attention in theoretical computer science , artificial intelligence , combinatorics and , more recently , statistical physics .
much of the work on random @xmath0-sat has focused on proving upper and lower bounds for @xmath23 , both for the smallest computationally hard case @xmath25 and for general @xmath0 . at this point
the existence of @xmath23 has not been established for any @xmath22 .
nevertheless , we will take the liberty of writing @xmath26 to denote that for all @xmath27 , @xmath28 is satisfiable ; analogously , we will write @xmath29 to denote that for all @xmath30 , @xmath28 is unsatisfiable . as we will see , an elementary counting argument yields @xmath31 for all @xmath0 .
lower bounds , on the other hand , have been exclusively algorithmic : to establish @xmath26 ones shows that for @xmath27 some specific algorithm finds a satisfying assignment with probability we will see that an extremely simple algorithm @xcite already yields @xmath32 .
we will also see that while more sophisticated algorithms improve this bound slightly , to date no algorithm is known to find a satisfying truth assignment ( even ) when @xmath33 , for any @xmath34 .
the threshold picture for hypergraph 2-colorability is completely analogous : for each @xmath8 , it is conjectured that there exists a constant @xmath35 such that @xmath36= \begin{cases } 1 & \mbox{if $ c < c_k$}\\ 0 & \mbox{if $ c > c_k$ } \enspace . \end{cases}\ ] ] the same counting argument here implies @xmath37 , while another simple algorithm yields @xmath38 .
again , no algorithm is known to improve this bound asymptotically , leaving a multiplicative gap of order @xmath39 between the upper and lower bound for this problem as well . in this paper
, we use the _ second moment method _ to show that random @xmath0-cnf formulas are satisfiable , and random @xmath0-uniform hypergraphs are 2-colorable , for density up to @xmath40 .
thus , we determine the threshold for random @xmath0-sat within a factor of two and the threshold for property b within a small additive constant .
recall that @xmath41 is unsatisfiable if @xmath42 .
our first main result is [ thm : ksat ] for all @xmath22 , @xmath43 is satisfiable if @xmath44 .
our second main result determines the property b threshold within an additive @xmath45 .
[ thm : hyp ] for all @xmath22 , @xmath46 is non-2-colorable if @xmath47 there exists a sequence @xmath48 such that for all @xmath22 , @xmath46 is 2-colorable if @xmath49 the upper bound in corresponds to the density for which the expected number of 2-colorings of @xmath50 is @xmath51 .
our main contribution is inequality which we prove using the second moment method .
in fact , our approach yields explicit bounds for the hypergraph 2-colorability threshold for each value of @xmath0 ( although ones that lack an attractive closed form ) .
we give the first few of these bounds in table [ tab : val ] .
we see that the gap between our upper and lower bounds converges to its limiting value of @xmath52 rather rapidly .
[ tab : val ] @xmath53 unlike the algorithmic lower bounds for random @xmath0-sat and hypergraph 2-colorability , our arguments are non - constructive : we establish that solutions exist for certain densities but do not offer any hint on how to find them .
we believe that abandoning the algorithmic approach for proving such lower bounds is natural and , perhaps , necessary . at a minimum ,
the algorithmic approach is limited to the small set of rather naive algorithms whose analysis is tractable using current techniques . perhaps more gravely , it could be that _ no _ polynomial algorithm can overcome the @xmath54 barrier .
determining whether this is true even for certain limited classes of algorithms , random walk algorithms , is a very interesting open problem .
in addition , by not seeking out some specific truth assignment , as algorithms do , the second moment method gives some first glimpses of the `` geometry '' of the set of solutions .
deciphering these first glimpses , getting clearer ones , and exploring potential interactions between the geometry of the set of solutions and computational hardness are great challenges lying ahead .
we note that recently , and independently , frieze and wormald @xcite applied the second moment method to random @xmath0-sat in the case where @xmath0 is a moderately growing function of @xmath11 .
specifically , they proved that when @xmath55 , @xmath9 is satisfiable if @xmath56 but unsatisfiable if @xmath57 , where @xmath58 and @xmath59 is such that @xmath60 .
their result follows by a direct application of the second moment method to the number of satisfying assignments of @xmath9 . as we will see shortly , while this approach gives a very sharp bound when @xmath61 , it fails for any fixed @xmath0 and indeed for any @xmath62 .
we also note that since this work first appeared @xcite , the line of attack we put forward has had several other successful applications .
specifically , in @xcite , the lower bound for the random @xmath0-sat threshold was improved to @xmath63 by building on the insights presented here . in @xcite
, the method was successfully extended to random max @xmath0-sat
@xcite , while in @xcite it was applied to random graph coloring .
we discuss these subsequent developments in the conclusions .
the version of the second moment method we will use is given by lemma [ lem : sec ] below and follows from a direct application of the cauchy - schwarz inequality ( see remark 3.1 in @xcite ) .
[ lem : sec ] for any non - negative random variable @xmath64 , @xmath65 \,\ge\ , \frac{\ex[x]^2}{\ex[x^2 ] } \enspace .\ ] ] it is natural to try to apply lemma [ lem : sec ] to random @xmath0-sat by letting @xmath64 be the number of satisfying truth assignments of @xmath9 .
unfortunately , as we will see , this naive " application of the second moment method fails rather dramatically : for all @xmath66 and every @xmath67 , @xmath68 > ( 1+\beta)^n \,\ex[x]^2 $ ] for some @xmath69 . as a result ,
the second moment method only gives an exponentially small lower bound on the probability of satisfiability .
the key step in overcoming this failure lies in realizing that we are free to apply the second moment method to any random variable @xmath64 such that @xmath70 implies that the formula is satisfiable .
in particular , we can let @xmath64 be the size of any _ subset _ of the set of satisfying assignments . by choosing this subset carefully ,
we can hope to significantly reduce the variance of @xmath64 relative to its expectation and use lemma [ lem : sec ] to prove that the subset is frequently non - empty .
indeed , we will establish the satisfiability of random @xmath0-cnf by focusing on those satisfying truth assignments _ whose complement is also satisfying_. in section [ sec : crapola ] we will give some intuition for why the number of such assignments has much smaller variance than the number of all satisfying assignments . for
now , we observe that considering only such satisfying assignments is equivalent to interpreting the random @xmath0-cnf formula @xmath9 as an instance of not - all - equal ( nae ) @xmath0-sat , where a truth assignment @xmath71 is a solution if and only if under @xmath71 every clause contains at least one satisfied literal _ and _ at least one unsatisfied literal . in other words , our lower bound for the @xmath0-sat threshold in theorem [ thm : ksat ] is , in fact , a lower bound for the nae @xmath0-sat threshold . indeed ,
for both random nae @xmath0-sat and random hypergraph 2-colorability we will apply lemma [ lem : sec ] naively , by letting @xmath64 be the number of solutions
. this will give theorem [ thm : hyp ] and the values in table [ tab : val ] for hypergraph 2-colorability and , as we will see , exactly the same bounds for random nae @xmath0-sat .
( indeed , the proof of theorem [ thm : hyp ] is a slight generalization of the proof for random nae @xmath0-sat . )
we will see that this success of the naive second moment is due to the symmetry inherent in both problems , to the fact that the complement of a solution is also a solution .
indeed , we feel that highlighting this role of symmetry and showing how it can be exploited even in asymmetric problems like @xmath0-sat is our main conceptual contribution .
exploiting these ideas in other constraint satisfaction problems that have a permutation group acting on the variables domain is an interesting area for further research . in section [ sec : back ] we give some background on random @xmath0-sat and random hypergraph 2-colorability . in section [ sec : crapola ]
we explain why the second moment method fails when applied to @xmath0-sat directly , and give some intuition for why counting only the nae - satisfying assignments rectifies the problem .
we also point out some connections to methods of statistical physics . in section [ sec :
ground ] we lay the groundwork for bounding the second moment for both nae @xmath0-sat and hypergraph 2-colorability by dealing with some probabilistic preliminaries , introducing a
laplace method " lemma for bounding certain sums , and outlining our strategy .
the actual bounding occurs in sections [ sec : nae_bound ] to [ sec : gleas ] .
specifically , in sections [ sec : nae_bound ] and [ sec : hyp ] we use the laplace lemma to reduce the second moment calculations for both random nae @xmath0-sat and random hypergraph 2-colorability to the maximization of a certain function @xmath72 on the unit interval , where @xmath72 is independent of @xmath11 .
we maximize @xmath72 in section [ sec : gleas ] and prove the laplace lemma in section [ sec : peak ] .
we conclude in section [ sec : conc ] by discussing some recent extensions of this work and proposing several open questions .
the mathematical investigation of random @xmath0-sat began with the work of franco and paull @xcite who , among other results , observed that @xmath43 is unsatisfiable if @xmath73 . to see this ,
let @xmath74 be the number of all possible @xmath0-clauses and let @xmath75 be the number of @xmath0-clauses consistent with a given truth assignment . since any fixed truth assignment is satisfying with probability @xmath76 , the expected number of satisfying truth assignments of @xmath43 is at most @xmath77^n = o(1)$ ] for @xmath78 . shortly afterwards , chao and franco @xcite complemented this result by proving that for all @xmath22 , if @xmath79 then the following linear - time algorithm , called unit clause ( uc ) , finds a satisfying truth assignment : if there exist unit clauses , pick one randomly and satisfy it ;
else pick a random unset variable and give it a random value .
note that since uc succeeds only ( rather than ) this does not imply a lower bound for @xmath23 .
the satisfiability threshold conjecture gained a great deal of popularity in the early 1990s and has received an increasing amount of attention since then .
the polynomial - time solvable case @xmath80 was settled early on : independently , chvtal and reed @xcite , fernandez de la vega @xcite , and goerdt @xcite proved that @xmath81 .
chvtal and reed @xcite , in addition to proving @xmath81 , gave the first lower bound for @xmath23 , strengthening the positive - probability result of chao and franco @xcite by analyzing the following refinement of uc , called short clause ( sc ) : if there exist unit clauses , pick one randomly and satisfy it ; else if there exist binary clauses , pick one randomly and satisfy a random literal in it ; else pick a random unset variable and give it a random value .
in @xcite , the authors showed that for all @xmath82 , sc finds a satisfying truth assignment for @xmath83 and raised the question of whether this lower bound for @xmath23 can be improved asymptotically .
a large fraction of the work on the satisfiability threshold conjecture since then has been devoted to the first computationally hard case , @xmath25 , and a long series of results @xcite has narrowed the potential range of @xmath84 .
currently this is pinned between @xmath85 by kaporis , kirousis and lalas @xcite and hajiaghayi and sorkin @xcite and @xmath86 by dubois and boufkhad @xcite .
upper bounds for @xmath84 come from probabilistic counting arguments , refining the above calculation of the expected number of satisfying assignments .
lower bounds , on the other hand , have come from analyzing progressively more sophisticated algorithms .
unfortunately , neither of these approaches helps narrow the asymptotic gap between the upper and lower bounds for @xmath23 .
the upper bounds only improve @xmath87 by a small additive constant ; the best algorithmic lower bound , due to frieze and suen @xcite , is @xmath88 where @xmath89 there are two more results that stand out in the study of random @xmath0-cnf formulas . in a breakthrough paper , friedgut @xcite proved the existence of a _ non - uniform _ satisfiability threshold , of a sequence @xmath90 around which the probability of satisfiability goes from 1 to 0 .
[ thm : frie ] for each @xmath91 , there exists a sequence @xmath90 such that for every @xmath92 , @xmath93 = \begin{cases } 1 & \mbox{if $ r = ( 1-\epsilon)\ , r_k(n)$}\\ 0 & \mbox{if $ r = ( 1+\epsilon)\ , r_k(n)$ } \enspace . \end{cases}\ ] ] in @xcite , chvtal and szemerdi established a seminal result in proof complexity , by extending the work of haken @xcite and urquhart @xcite to random formulas .
specifically , they proved that for all @xmath22 , if @xmath73 then @xmath94 is unsatisfiable but every resolution proof of its unsatisfiability contains at least @xmath95 clauses , for some @xmath96 . in @xcite , achlioptas , beame and
molloy extended the main result of @xcite to random cnf formulas that also contain as this is relevant for the behavior of davis - putnam ( dpll ) algorithms on random @xmath0-cnf .
( dpll algorithms proceed by setting variables sequentially , according to some heuristic , and backtracking whenever a contradiction is reached . ) by combining the results in the present paper with the results in @xcite , it was recently shown @xcite that a number of dpll algorithms require exponential time _ significantly below _ the satisfiability threshold , for provably satisfiable random @xmath0-cnf formulas .
finally , we note that if one chooses to live unencumbered by the burden of mathematical proof , powerful non - rigorous techniques of statistical physics , such as the `` replica method '' , become available .
indeed , several claims based on the replica method have been subsequently established rigorously , so it is frequently ( but definitely not always ) correct . using this technique , monasson and zecchina
@xcite predicted @xmath97 . like most arguments based on the replica method ,
their argument is mathematically sophisticated but far from rigorous . in particular
, they argue that as @xmath0 grows large , the so - called _ annealed approximation _ should apply .
this creates an analogy with the second moment method which we discuss in section [ sec : replica ] .
while bernstein @xcite originally raised the 2-colorability question for certain classes of infinite set families , erds popularized the finite version of the problem @xcite and the hypergraph representation . recall that a 2-uniform hypergraph , a graph , is 2-colorable if and only if it has no odd cycle . in a random graph with @xmath98 edges this occurs with constant probability if and only if @xmath99 ( see @xcite for more on the evolution of cycles in random graphs ) . for all @xmath100 , on the other hand , hypergraph 2-colorability is np - complete @xcite and determining the 2-colorability threshold @xmath35 for @xmath0-uniform hypergraphs @xmath50 remains open .
analogously to random @xmath0-sat , we will take the liberty of writing @xmath101 if @xmath50 is 2-colorable for all @xmath102 , and @xmath103 if @xmath50 is non-2-colorable for all @xmath104 .
alon and spencer @xcite were the first to give bounds on the potential value of @xmath35 .
specifically , they observed that , analogously to random @xmath0-sat , the expected number of 2-colorings of @xmath50 is at most @xmath105^n$ ] and concluded that @xmath50 is non-@xmath0-colorable if @xmath106 .
more importantly , by employing the lovsz local lemma , they proved that @xmath50 is 2-colorable if @xmath107 . regarding the upper bound , it is easy to see that , in fact , @xmath108 if @xmath109 and this yields the upper bound of theorem [ thm : hyp ] . moreover , the techniques of @xcite can be used to improve this bound further to @xmath110 , where @xmath48 .
the lower bound of @xcite was improved by achlioptas , kim , krivelevich and tetali @xcite motivated by the analogies drawn in @xcite between hypergraph 2-colorability and earlier work @xcite for random @xmath0-sat .
specifically , it was shown in @xcite that a simple , linear - time algorithm finds a 2-coloring of @xmath50 for @xmath111 , implying @xmath112 .
these were the best bounds for @xmath35 prior to theorem [ thm : hyp ] of the present paper .
finally , we note that friedgut s result @xcite applies to hypergraph 2-colorability as well , [ thm : frie_hyp ] for each @xmath22 , there exists a sequence @xmath113 such that for every @xmath92 , @xmath114 = \begin{cases } 1 & \mbox{if $ c = ( 1-\epsilon)\ , c_k(n)$}\\ 0 & \mbox{if $ c = ( 1+\epsilon)\ , c_k(n)$ } \enspace . \end{cases}\ ] ]
in the rest of the paper it will be convenient to work with a model of random formulas that differs slightly from @xmath9 . specifically , to generate a random @xmath0-cnf formula on @xmath11 variables with @xmath12 clauses we simply generate a string of @xmath115 independent random literals , each such literal drawn uniformly from among all @xmath116 possible ones .
note that this is equivalent to selecting , with replacement , @xmath12 clauses from among all possible @xmath117 ordered @xmath0-clauses .
this choice of distribution for @xmath0-cnf formulas will simplify our calculations significantly . as we will see in section [ sec : gene ] , the derived results can be easily transferred to all other standard models for random @xmath0-cnf formulas . for any formula @xmath7 , given truth assignments
@xmath118 we will write @xmath119 to denote that _ all _ of @xmath120 satisfy @xmath7 .
let @xmath121 denote the number of satisfying assignments of a formula @xmath7 .
then , for a @xmath0-cnf formula with random clauses @xmath122 we have @xmath123 = \ex\left[\sum_{\sigma } \ones_{\sigma \models f}\right ] = \sum_{\sigma}\ex\left[\prod_{c_i}\ones_{\sigma \models c_i } \right ] = \sum_{\sigma}\prod_{c_i } \ex[\ones_{\sigma \models c_i } ] = 2^n ( 1 - 2^{-k})^{m } \enspace , \ ] ] since clauses are drawn independently and the probability that @xmath71 satisfies the @xmath124th random clause , @xmath125 $ ] , is @xmath126 for every @xmath71 and @xmath124 .
similarly , for @xmath68 $ ] we have @xmath127 = \ex\left[\left(\sum_{\sigma } \ones_{\sigma \models f}\right)^2\right ] = \ex\left[\sum_{\sigma,\tau } \ones_{\sigma,\tau \models f}\right ] = \sum_{\sigma,\tau}\ex\left[\prod_{c_i}\ones_{\sigma,\tau \models c_i } \right ] = \sum_{\sigma,\tau}\prod_{c_i } \ex[\ones_{\sigma,\tau \models c_i } ] \enspace .\ ] ] we claim that @xmath128 $ ] , the probability that a fixed pair of truth assignments @xmath129 satisfy the @xmath124th random clause , depends only on the number of variables @xmath130 to which @xmath71 and @xmath131 assign the same value .
specifically , if the overlap is @xmath132 , we claim that this probability is @xmath133 our claim follows by inclusion - exclusion and observing that if @xmath134 is not satisfied by @xmath71 , the only way for it to also not be satisfied by @xmath131 is for all @xmath0 variables in @xmath134 to lie in the overlap of @xmath71 and @xmath131 .
thus , @xmath135 quantifies the correlation between the events that @xmath71 and @xmath131 are satisfying as a function of their overlap .
in particular , observe that truth assignments with overlap @xmath136 are uncorrelated since @xmath137 ^ 2 $ ] .
since the number of ordered pairs of assignments with overlap @xmath130 is @xmath138 we thus have @xmath139 = 2^n \sum_{z=0}^n \binom{n}{z } \,f_s(z / n)^{m } \enspace .\ ] ] writing @xmath140 and using the approximation @xmath141 we see that @xmath139 \geq 2^n \left(\max_{0 \leq \a \leq 1 } \left[\frac{f_s(\a)^r}{\a^\a(1-\a)^{1-\a}}\right ] \right)^n \times { \mathrm { poly}}(n ) \equiv \left(\max_{0 \leq \a \leq 1 } \lambda_s(\a)\right)^n \times { \mathrm { poly}}(n ) \enspace .\ ] ] at the same time observe that @xmath142 ^ 2 = \left(2^n ( 1 - 2^{-k})^{rn}\right)^2 = \left(4f_s(1/2)^r\right)^n=\lambda_s(1/2)^n$ ] .
therefore , if there exists some @xmath143 $ ] such that @xmath144 then the second moment is exponentially greater than the square of the expectation and we only get an exponentially small lower bound for @xmath145 $ ]
. put differently , unless the dominant contribution to @xmath68 $ ] comes from uncorrelated " pairs of satisfying assignments , pairs with overlap @xmath136 , the second moment method fails . with these observations in mind , in fig .
1 we plot @xmath146 for @xmath147 and different values of @xmath148 .
we see that , unfortunately , for all values of @xmath148 shown @xmath149 is maximized at some @xmath150 .
if we look closely into the two factors comprising @xmath149 , the reason for the failure of the second moment method becomes apparent : while the entropic factor @xmath151 is symmetric around @xmath52 , the correlation function @xmath135 is strictly increasing in @xmath152 $ ] .
therefore , the derivative of @xmath149 is never 0 at 1/2 , instead becoming 0 at some @xmath150 where the benefit of positive correlation balances with the cost of decreased entropy .
( indeed , this is true for all @xmath62 and constant @xmath153 . )
[ fig_sat ] @xmath147 , @xmath154 ( top to bottom ) let us now repeat the above analysis but with @xmath121 being the number of nae - satisfying truth assignments of a formula @xmath7 . recall that @xmath71 is a nae - satisfying assignment iff under @xmath71 every clause has at least one satisfied literal _ and _ at least one unsatisfied literal .
thus , for a @xmath0-cnf formula with random clauses @xmath122 , proceeding as in , we get @xmath155 = 2^n ( 1 - 2^{1-k})^{m } \enspace , \ ] ] since the probability that @xmath71 nae - satisfies the @xmath124th random clause is @xmath156 for every @xmath71 and @xmath124 .
regarding the second moment , proceeding exactly as in , we write @xmath68 $ ] as a sum over the @xmath157 ordered pairs of assignments of the probability that both assignments are nae - satisfying . as for @xmath0-sat , for any fixed pair this probability depends only on the overlap .
the only change is that if @xmath129 agree on @xmath140 variables then the probability they both nae - satisfy a random clause @xmath134 is @xmath158 & = & 1- 2^{2-k}+2^{1-k}\left(\a^k+(1-\a)^k\right ) \nonumber \\ & \equiv & f_n(\alpha)\enspace \label{eq : fn } \enspace .\end{aligned}\ ] ] again , this claim follows from inclusion - exclusion and observing that for @xmath129 to both nae - violate @xmath134 , the variables of @xmath134 must either all be in the overlap of @xmath71 and @xmath131 or all be in their non - overlap . applying stirling s approximation for the factorial again and observing that the sum defining @xmath68 $ ] has only a polynomial number of terms , we now get @xmath159 \leq 2^n \left(\max_{0 \leq \a \leq 1 } \left[\frac{f_n(\a)^r}{\a^\a(1-\a)^{1-\a}}\right ] \right)^n \times { \mathrm { poly}}(n ) \equiv \left(\max_{0 \leq \a \leq 1 } \lambda_n(\a)\right)^n \times { \mathrm { poly}}(n ) \enspace .\ ] ] as before , it is easy to see that @xmath142 ^ 2 = \lambda_n(1/2)^n$ ] . therefore , if @xmath160 for every @xmath161 then implies that the ratio between @xmath68 $ ] and @xmath142 ^ 2 $ ] is at most polynomial in @xmath11 . indeed , with a more careful analysis of the interplay between the summation and stirling s approximation , we will later show that whenever @xmath162 is a global maximum , the ratio @xmath68/\ex[x]^2 $ ] is bounded by a constant , implying that nae - satisfiability holds so , all in all , again we hope that the dominant contribution to @xmath68 $ ] comes from pairs of assignments with overlap @xmath136 .
the crucial difference is that now the correlation function @xmath163 is _ symmetric _ around @xmath52 and , hence , so is @xmath164 . as a result
, the entropy - correlation product @xmath164 always has a local extremum at @xmath52 . moreover , since the entropic term is always maximized at @xmath165 and is independent of @xmath148 , for sufficiently small @xmath148 this extremum is a global maximum . with these considerations in mind , in fig . 2 we plot @xmath166 for @xmath147 and various values of @xmath148 .
let us start with the picture on the left , where @xmath148 increases from 8 to 12 as we go from top to bottom .
for @xmath167 we see that indeed @xmath164 has a global maximum at @xmath52 and the second moment method succeeds . for the cases @xmath168 , on the other hand , we see that @xmath162 is actually a global minimum .
in fact , we see that @xmath169 , implying that @xmath142 ^ 2 = \lambda_n(1/2)^n = o(1)$ ] and so there are no nae - satisfying assignments for such @xmath148 .
it is worth noting that for @xmath170 , even though @xmath171 , the second moment is exponentially large ( since @xmath172 near @xmath173 and @xmath174 ) .
[ fig_nae ] @xmath147 , @xmath175 ( top to bottom ) @xmath147 , @xmath176 the most interesting case is @xmath177 . here @xmath178 is a local maximum and greater than 1 , but the two global maxima occur at @xmath179 and @xmath180 where the function equals @xmath181 ... as a result , again , the second moment method only gives an exponentially small lower bound on @xmath182 $ ] .
note that this is in spite of the fact that @xmath142 $ ] is now exponentially large .
indeed , the largest value for which the second moment succeeds for @xmath147 is @xmath183 when the two side peaks reach the same height as the peak at @xmath52 ( see the plot on the right in fig . 2 ) .
so , the situation can be summarized as follows .
by requiring that we only count nae - satisfying truth assignments we make it , roughly , twice as hard to satisfy each clause .
this manifests itself in the additional factor of 2 in the middle term of @xmath163 compared to @xmath135 . on the other hand , now , the third term of @xmath184 , capturing
joint " behavior , is symmetric around @xmath52 , making @xmath185 itself symmetric around @xmath52 .
this enables the second moment method which , indeed , only breaks down when the density gets within an additive constant of the upper bound for the nae @xmath0-sat threshold .
given a truth assignment @xmath71 and an arbitrary cnf formula @xmath7 , let @xmath186 denote the total number of literal _ occurrences _ in @xmath7 satisfied by @xmath71 .
so , for example , @xmath187 is maximized by those truth assignments that assign every variable its `` majority '' value . with this definition at hand , a potential explanation of how symmetry reduces the variance is suggested by considering the following trivial refinement of our generative model : first i ) draw @xmath115 i.i.d .
uniformly random literals just as before and then ii ) partition the drawn literals randomly into @xmath0-clauses ( rather than assuming that the first @xmath0 literals form the first clause , the next @xmath0 the second , etc . ) .
in particular , imagine that we have just finished performing the first generative step above and we are about to perform the second .
observe that at this point the value of @xmath187 has already been determined for every @xmath188 .
moreover , for each fixed @xmath71 the conditional probability of yielding a satisfying assignment corresponds to a balls - in - bins experiment : distribute @xmath189 balls in @xmath12 bins , each with capacity @xmath0 , so that every bin receives at least one ball .
it is clear that those truth assignments for which @xmath187 is large at the end of the first step have a big advantage in the second .
to get an idea of what @xmath187 typically looks like on @xmath190 we begin by observing that the number of occurrences of a fixed literal @xmath191 , @xmath192 , is distributed as @xmath193 .
thus , @xmath194 = o(1)$ ] and , moreover , the random variables @xmath192 are very weakly correlated . as a result ,
at the end of the first step , @xmath187 is typically a very smooth function on @xmath190 , attaining a maximum value at the subcube of majority vote assignments and gradually decreasing away from them .
thus , at the end of the first step the more promising " truth assignments are correlated : in satisfying many literal occurrences ( thus increasing their odds for the second step ) , they tend to overlap with each other ( and the majority assignment ) at more than half the variables .
in contrast , if we focus on nae - satisfying assignments , at the end of the first step the most promising assignments @xmath71 are those for which @xmath189 is very close to its average value @xmath195 .
so , when the problem is symmetric , the typical case becomes the most favorable case and the clustering around truth assignments that satisfy many literal occurrences disappears .
if indeed `` populism '' , the tendency of each variable to assume its majority value in the formula , is the main source of correlations in random @xmath0-sat , then the second moment method is a good candidate for @xmath0-cnf models which do not encourage this tendency .
for example , one such model is _ regular _ random @xmath0-sat , in which every literal occurs exactly the same number of times .
such formulas can be analyzed using a model analogous to the configuration model of random graphs , by taking precisely @xmath196 copies of each literal and partitioning the resulting @xmath197 copies into clauses randomly ( exactly as in the second step of our two - step model for random @xmath0-sat ) .
a key quantity in statistical physics is the _ overlap distribution _ between configurations of minimum energy , known as ground states .
when a constraint satisfaction problem is satisfiable , ground states correspond to solutions , such as satisfying assignments , 2-colorings , and so on . in the case of random , the overlap distribution is the probability @xmath198 that a random pair of satisfying assignments have overlap @xmath199 .
this overlap distribution gives an intriguing interpretation of our results . in writing @xmath68
$ ] as a sum of contributions from pairs of assignments with different overlaps , we have in fact calculated the average of @xmath198 over all formulas , weighted by the number of pairs of satisfying assignments of each one .
physicists call this weighted average the `` annealed approximation '' of @xmath198 , and denote it @xmath200 .
it is worth pointing out that , while the annealed approximation clearly overemphasizes formulas with more satisfying assignments , monasson and zecchina conjectured in @xcite , based on the replica method , that it becomes asymptotically tight as @xmath201 . on a more rigorous footing
, it is easy to see that in our case @xmath200 is proportional to @xmath202 .
therefore , whenever @xmath185 is peaked at @xmath52 , @xmath200 is tightly peaked around @xmath52 , since @xmath202 vanishes for all other values of @xmath203 as @xmath204 .
this is precisely what we prove occurs in random for densities up to @xmath205 . in other words , for densities
almost all the way to the random nae @xmath0-sat threshold , in the annealed approximation , the nae - satisfying assignments are scattered throughout the hypercube _ as if they were independent . _
note that even if @xmath198 is concentrated around 1/2 ( rather than just @xmath200 ) this still allows for a typical geometry where there are exponentially many , exponentially large clusters , each centered at a random assignment .
indeed , this is precisely the picture suggested by some very recent , ground - breaking work of mezard , parisi , and zecchina @xcite , based on non - rigorous techniques of statistical physics .
if this is indeed the true picture , establishing it rigorously would require considerations much more refined than the second moment of the number of solutions . more generally , getting a better understanding of the typical geometry and its potential implications for algorithms appears to us a very challenging and very important open problem .
given a set @xmath3 of @xmath11 boolean variables , let @xmath206 denote the set of all proper _ @xmath0-clauses _ on @xmath3 , the set of all @xmath207 disjunctions of @xmath0 literals involving distinct variables .
similarly , given a set @xmath3 of @xmath11 vertices , let @xmath208 be the set of all @xmath209 @xmath0-subsets of @xmath3 . as we saw , a random @xmath0-cnf formula @xmath9 is formed by selecting uniformly a random @xmath12-subset of @xmath210 , while a random @xmath0-uniform hypergraph @xmath13 is formed by selecting uniformly a random @xmath12-subset of @xmath211 .
while @xmath9 and @xmath13 are perhaps the most natural models for generating random @xmath0-cnf formulas and random @xmath0-uniform hypergraphs , respectively , there are a number of slight variations of each model .
those are largely motivated by amenability to certain calculations . to simplify the discussion we focus on models for random formulas in the rest of this subsection .
all our comments transfer readily to models for random hypergraphs .
for example , it is fairly common to consider the clauses as ordered @xmath0-tuples ( rather than as @xmath0-sets ) and/or to allow replacement in sampling the set @xmath210 .
clearly , for properties such as satisfiability the issue of ordering is irrelevant .
moreover , as long as @xmath212 , essentially the same is true for the issue of replacement . to see that ,
observe that the number of repeated clauses is @xmath213 and the subset of @xmath214 distinct clauses is uniformly random .
thus , if a monotone decreasing property ( such as satisfiability ) holds with probability @xmath215 for a given @xmath216 when replacement is allowed , it holds with probability @xmath217 for all @xmath27 when replacement is not allowed .
the issue of selecting the literals of each clause with replacement ( which might result in some improper " clauses ) is completely analogous .
that is , the probability that a variable appears more than once in a given clause is at most @xmath218 and hence there are @xmath219 improper clauses .
finally , we note that by standard techniques our results also transfer to the @xmath220 model where every clause appears independently of all others with probability @xmath215 . for that it suffices to set @xmath215 such that @xmath221 .
our plan is to consider random @xmath0-cnf formulas formed by generating @xmath115 i.i.d .
random literals , where @xmath222 , and proving that if @xmath121 is the number of nae - satisfying assignments then : [ lem : naepos ] for all @xmath223 , @xmath224 and @xmath225 , there exists some constant @xmath226 such that @xmath227 }
< c \times { \ex[x]^2 } \enspace .\ ] ] by lemma [ lem : sec ] and our discussion in section [ sec : gene ] , this implies that @xmath228 is nae - satisfiable since a nae - satisfiable formula is also satisfiable , we have established that @xmath41 is satisfiable for all @xmath148 as in lemma [ lem : naepos ] . to boost this to a high probability result , thus establishing theorem [ thm : ksat ] , we employ the following immediate corollary of theorem [ thm : frie ] .
[ cor : boost_nae ] if @xmath229 is satisfiable then @xmath41 is satisfiable for all @xmath230 .
friedgut s arguments @xcite apply equally well to nae @xmath0-sat , implying that @xmath41 is nae - satisfiable for @xmath148 as in lemma [ lem : naepos ] .
thus , lemma [ lem : naepos ] readily yields below , while comes from noting that the expected number of nae - satisfying assignments is @xmath231^n$ ] .
( similarly to hypergraphs , the techniques of @xcite can be used to improve the bound in to @xmath110 , where @xmath48 . )
indeed , we will see that the proof of theorem [ thm : naeksat ] will yield theorem [ thm : hyp ] for random hypergraphs with little additional effort .
[ thm : naeksat ] for all @xmath22 , @xmath43 is non - nae - satisfiable if @xmath232 there exists a sequence @xmath48 such that for all @xmath22 , @xmath43 is nae - satisfiable if @xmath233 as we saw in section [ july4 ] , the second moment of the number of nae - satisfying assignments is @xmath234 a slightly more complicated sum will occur when we bound the second moment of the number of 2-colorings . to bound both sums we will use the following lemma which we prove in section [ sec : peak ] . [
lem : peak ] * [ laplace lemma ] * let @xmath235 be a positive , twice - differentiable function on @xmath152 $ ] and let @xmath236 be a fixed integer .
let @xmath237 and let @xmath238 letting @xmath239 , define @xmath72 on @xmath152 $ ] as @xmath240 if there exists @xmath241 such that @xmath242 for all @xmath243 , and @xmath244 , then there exists a constant @xmath245 such that @xmath246
recall that if @xmath64 is the number of nae - assignments , then @xmath155 = 2^n ( 1 - 2^{1-k})^{rn}\ ] ] and @xmath247 = 2^n \sum_{z=0}^n \binom{n}{z } \,f_n(z / n)^{rn } \enspace , \ ] ] where @xmath248 to bound the sum in we apply lemma [ lem : peak ] with @xmath249 and @xmath250 .
thus , @xmath251 where @xmath252 to show that lemma [ lem : peak ] applies , we will prove in section [ sec : gleas ] that [ gleas ] for every @xmath223 , there exists @xmath253 such that for all @xmath254 , if @xmath255 then @xmath256 for all @xmath257 , and @xmath258 .
therefore , for all @xmath148 , @xmath0 , and @xmath259 as in lemma [ gleas ] , there exists a constant @xmath260 such that @xmath227 } < c \times { 2^n g_r(1/2)^n } \enspace .\ ] ] since @xmath142 ^ 2 = 2^n g_r(1/2)^n$ ] we get that for all @xmath261 as in lemma [ gleas ] @xmath227 } < c \times { \ex[x]^2 } \enspace .\ ] ]
just as for nae @xmath0-sat , it will be easier to work with the model in which generating a random hypergraph corresponds to generating @xmath115 random vertices , each such vertex chosen uniformly at random with replacement , and letting the first @xmath0 vertices form the first hyperedge etc . in @xcite we proved of theorem [ thm : hyp ] by letting @xmath64
be the set of all 2-colorings and using a convexity argument to show that @xmath68 $ ] is dominated by the contribution of _ balanced _ colorings , colorings with an equal number of black and white vertices . here
we follow a simpler approach suggested by david karger ; namely , we _ define _ @xmath64 to be the number of balanced 2-colorings .
we emphasize that , while technically convenient , the restriction to balanced 2-colorings is not essential for the second moment method to succeed on hypergraph 2-colorability , one has @xmath68 = o(\ex[x]^2)$ ] even if @xmath64 is the number of all 2-colorings . of course , in order for balanced colorings to exist @xmath11 must be even and we will assume that in our calculations below . to get theorem [ thm : hyp ] for all sufficiently large @xmath11 , we observe that if for a given @xmath262 , @xmath263 is 2-colorable then for all @xmath264 , @xmath50 is 2-colorable since deleting a random vertex of @xmath265 removes @xmath219 edges . with this in mind , in the following we let @xmath64 be the number of balanced 2-colorings and assume that @xmath11 is even . since the vertices in each hyperedge are chosen uniformly with replacement , then
the probability that a random hyperedge is bichromatic in a fixed balanced partition is @xmath156 .
since there are @xmath266 such partitions and the @xmath12 hyperedges are drawn independently , we have @xmath267 = \binom{n}{n/2 } \,\left(1 - 2^{1-k}\right)^{m } \enspace .\ ] ] to calculate the second moment , as we did for [ nae ] @xmath0-sat , we write @xmath68 $ ] as a sum over all pairs of balanced partitions . in order to estimate this sum
we first observe that if two balanced partitions @xmath71 and @xmath131 have exactly @xmath130 black vertices in common , then they must also have exactly @xmath130 white vertices in common .
thus @xmath71 and @xmath131 define four groups of vertices : @xmath130 that are black in both , @xmath130 that are white in both , @xmath268 that are black in @xmath71 and white in @xmath131 , and @xmath268 that are white in @xmath71 and black in @xmath131 . clearly , a random hyperedge is monochromatic in both @xmath71 and @xmath131 iff all its vertices fall into the same group .
since the vertices of each hyperedge are chosen uniformly with replacement , this probability is @xmath269 \enspace .\ ] ] thus , by inclusion - exclusion , the probability that a random hyperedge is bichromatic in both @xmath71 and @xmath131 is @xmath270 = f_n(2z / n)\ ] ] where @xmath271 is the function we defined for nae @xmath0-sat in .
moreover , observe that the number of pairs of partitions with such overlap is @xmath272 since hyperedges are drawn independently and with replacement , by summing over @xmath130 we thus get @xmath139 = \binom{n}{n/2 } \,\sum_{z=0}^{n/2 } \binom{n/2}{z}^{\!2 } f_n(2z / n)^{cn } \enspace .\ ] ] to bound this sum we apply lemma [ lem : peak ] with @xmath273 and @xmath274 .
felicitously , we find ourselves maximizing a function @xmath275 which , if we replace @xmath276 with @xmath148 , is exactly the same function @xmath277 we defined in for nae @xmath0-sat .
thus , setting @xmath278 where @xmath279 and @xmath259 are as in lemma [ gleas ] , @xmath275 is maximized at @xmath165 with @xmath280 , and lemma [ lem : peak ] implies that there exists a constant @xmath281 such that @xmath227 } < c \ , { n^{-1/2 } \ , \binom{n}{n/2 } \,g_c(1/2)^n } \enspace .\ ] ] we now bound @xmath142 $ ] from below using stirling s approximation and get @xmath282}{\ex[x]^2 } < c
\times \frac{n^{-1/2 } \binom{n}{n/2 } \,g_c(1/2)^n}{{\binom{n}{n/2}}^2 \,(1 - 2^{1-k})^{2rn } } = c \times \frac{n^{-1/2}\,2^n}{\binom{n}{n/2 } } \to c \times \sqrt{\frac{\pi}{2 } } \enspace .\ ] ] to complete the proof , analogously to [ nae ] @xmath0-sat , we use the following boosting " corollary of theorem [ thm : frie_hyp ] .
[ cor : boost_hyp ] if @xmath283 is 2-colorable then @xmath50 is 2-colorable for all @xmath102 .
we need to prove @xmath284 and @xmath256 for all @xmath285 .
as @xmath277 is symmetric around @xmath52 , we can restrict to @xmath286 $ ] .
we divide @xmath287 $ ] into two parts and handle them with two separate lemmata . the first lemma deals with
@xmath288 $ ] and also establishes that @xmath284 .
[ lem : nearhalf ] let @xmath288 $ ] . for all @xmath289 , if @xmath290 then @xmath256 and @xmath258 . the second lemma deals with @xmath291 $ ] .
[ lem : far ] let @xmath291 $ ] . for every @xmath223 and all @xmath224 , if @xmath292 then @xmath256 . combining lemmata [
lem : nearhalf ] and [ lem : far ] we see that for every @xmath223 and @xmath293 , if @xmath294 then @xmath256 for all @xmath285 and @xmath295 , establishing lemma [ gleas ] .
we prove lemmata [ lem : nearhalf ] and [ lem : far ] below .
the reader should keep in mind that we have made no attempt to optimize the value of @xmath296 in lemma [ lem : far ] , aiming instead for proof simplicity . for the lower bounds presented in table [ tab : val ] we computed numerically , for each @xmath0 , the largest value of @xmath148 for which the conclusions of lemma [ gleas ] hold . in each case , the condition @xmath297 was satisfied with room to spare , while establishing @xmath298 for all @xmath257 was greatly simplified by the fact that @xmath72 always has no more than three local extrema in @xmath152 $ ] .
* proof of lemma [ lem : nearhalf ] .
* we will first prove that for @xmath299 , @xmath277 is strictly decreasing in @xmath300 $ ] , thus establishing @xmath301 .
since @xmath277 is positive , to do this it suffices to prove that @xmath302 in this interval .
in fact , since @xmath303 at @xmath165 , it will suffice to prove that for @xmath304 $ ] we have @xmath305 .
now , @xmath306 to show that the r.h.s .
of is negative we first note that for @xmath307 and @xmath308 , @xmath309 is monotonically increasing .
therefore , @xmath310 .
moreover , for all @xmath203 , @xmath311 . therefore , since @xmath312 and @xmath313 , it suffices to observe that for all @xmath314 , @xmath315 finally , recalling that @xmath316 and using @xmath317 we see that @xmath318 since @xmath319.@xmath320 * proof of lemma [ lem : far ] .
* by the definition of @xmath277 we see that @xmath256 if and only if @xmath321 letting @xmath322 denote the entropy function , we see that holds as long as @xmath323 where @xmath324 observe now that for @xmath308 , @xmath184 is strictly increasing in @xmath287 $ ] , so @xmath325 .
moreover , for any @xmath326 @xmath327 since @xmath328 and @xmath329 , we thus see that it suffices to have @xmath330 now observe that for any @xmath331 and @xmath332 , @xmath333 since @xmath334 we can set @xmath335 , yielding @xmath336 since @xmath337 , we find that holds as long as @xmath338 where @xmath339 we are thus left to minimize @xmath235 in @xmath340 $ ] .
since @xmath235 is differentiable its minima can only occur at @xmath341 or @xmath174 , or where @xmath342 .
the derivative of @xmath235 is @xmath343 \enspace .
\nonumber \label{eq : phiprime}\ ] ] note now that for all @xmath344 @xmath345 is positively infinite . at the same time , @xmath346 is negative for @xmath347 .
therefore , @xmath235 is minimized in the interior of @xmath340 $ ] for all @xmath348 .
setting @xmath349 to zero gives @xmath350 by `` bootstrapping ''
we derive a tightening series of lower bounds on the solution for the l.h.s . of for @xmath351 .
note first that we have an easy upper bound , @xmath352 at the same time , if @xmath353 then @xmath354 , implying @xmath355 if we write @xmath356 then becomes @xmath357 by inspection , if @xmath358 the r.h.s .
of is greater than the l.h.s . for all @xmath359 , yielding a contradiction .
therefore , @xmath360 for all @xmath353 .
since @xmath361 for @xmath359 , we see that for @xmath362 , implies @xmath363 finally , observe that implies that as @xmath0 increases the denominator of approaches @xmath174 . to bootstrap
, we note that since @xmath334 we have @xmath364 where relies on and .
moreover , @xmath334 implies @xmath365 .
thus , by using and the fact @xmath366 for all @xmath367 , gives for @xmath22 , @xmath368 for @xmath369 , @xmath370 .
thus , by , we have @xmath371 .
this , in turn , implies @xmath372 and so , by and , we have for @xmath359 @xmath373 plugging into to bootstrap again , we get that for @xmath22 @xmath374 since @xmath375 for @xmath376 and @xmath377 for @xmath378 , we see that for such @xmath0 @xmath379 plugging into the fact @xmath380 we get @xmath381 . using that @xmath382 for @xmath383 , we get the closely matching upper bound , @xmath384 thus , we see that for @xmath369 , @xmath235 is minimized at an @xmath385 which is within @xmath386 of @xmath126 , where @xmath387 .
let @xmath388 be the interval @xmath389 $ ] .
clearly the minimum of @xmath235 is at least @xmath390 .
it is easy to see from that if @xmath391 then @xmath392 .
now , a simple calculation using that @xmath393 for @xmath394 gives @xmath395 therefore , @xmath396 finally , recall that holds as long as @xmath397 , @xmath398 clearly , we can take @xmath399 so that for all @xmath400 the error term @xmath401 is smaller than any @xmath223 . @xmath320
the idea behind lemma [ lem : peak ] is that sums of this type are dominated by the contribution of @xmath402 terms around the maximum term .
the proof amounts to replacing the sum by an integral and using the laplace method for asymptotic integrals @xcite .
we start by establishing two upper bounds for the terms of @xmath403 , one crude and one sharp . for the sharp bound we will use the following form of stirling s approximation , valid for all @xmath404 : @xmath405 recall that the @xmath130th term of @xmath403 is @xmath406 , where @xmath407 and @xmath408 . fix any @xmath409 and suppose that @xmath410 where @xmath411 $ ] .
then yields@xmath412 where @xmath413 .
in addition to , valid for @xmath414 $ ] , we will also use a cruder bound , valid for all @xmath415 .
namely , using the upper bound of for @xmath416 and the lower bound @xmath417 for @xmath418 and @xmath419 ( where we take @xmath239 ) we get ( since @xmath420 ) @xmath421 recall now that @xmath422 for all @xmath243 . if @xmath423 denotes the interval @xmath424 $ ] then for every @xmath223 , there exists a constant @xmath425 such that @xmath426 for all @xmath427 .
let @xmath428 and @xmath429 , and let @xmath430 we use to bound the terms in @xmath431 and to bound the remaining terms of @xmath403 . since @xmath432 , and since @xmath433 for any @xmath434 , we see that for every @xmath223@xmath435 say that a twice - differentiable function @xmath436 is _ unimodal _ on an interval @xmath437 $ ] if @xmath438 has a unique zero @xmath439 $ ] with @xmath440 , and furthermore @xmath441 . since @xmath442 for all @xmath443 and @xmath444 , we can take @xmath259 small enough so that @xmath72 is unimodal on @xmath423 .
this implies that @xmath445 is also unimodal on @xmath446 and , for @xmath447 , that @xmath448 is unimodal also .
since @xmath448 is unimodal , @xmath449 we evaluate this last integral using lemma [ lem : debruijn ] , the laplace method for asymptotic integrals .
[ lem : debruijn]@xcite .
let @xmath450 be unimodal on @xmath437 $ ] where @xmath276 is the unique zero of @xmath451 in @xmath437 $ ] .
then @xmath452 applying lemma [ lem : debruijn ] to with @xmath453 and @xmath454 , we see that @xmath455 , where @xmath456.@xmath320 lemma [ lem : peak ] has the following obvious corollary , which is useful for a variety of second moment calculations .
let @xmath403 and @xmath457 be defined as in lemma [ lem : peak ] .
if there exists @xmath241 with @xmath458 and a constant @xmath459 such that @xmath460 for all @xmath203 , then there exists a constant @xmath461 such that @xmath246
before this work , lower bounds on the thresholds of random constraint satisfaction problems were largely derived by analyzing very simple heuristics . here , instead , we derive such bounds by applying the second moment method to the number of solutions .
in particular , for random nae @xmath0-sat and random hypergraph 2-colorability we determine the location of the threshold within a small additive constant for all @xmath0 . as a corollary
, we establish that the asymptotic order of the random @xmath0-sat threshold is @xmath2 answering a long - standing open question .
since this work first appeared @xcite , our methods have been extended and applied to other problems . for random @xmath0-sat , achlioptas and peres
@xcite confirmed our suspicion ( see section [ sec : boost ] ) that the main source of correlations in random @xmath0-sat is the `` populist '' tendency of satisfying assignments towards the majority vote assignment . by considering a carefully constructed random variable which focuses on balanced solutions ,
on satisfying assignments that satisfy roughly half of all literal occurrences , they showed @xmath462 , establishing @xmath463 . in @xcite , achlioptas , naor and peres extended the approach of balanced solutions to max @xmath0-sat .
let us say that a @xmath0-cnf formula is @xmath215-satisfiable if there exists a truth assignment which satisfies at least @xmath464 of all clauses ; note that every @xmath0-cnf is 0-satisfiable .
for @xmath465 $ ] let @xmath466 denote the threshold for @xmath43 to be @xmath215-satisfiable ( so that @xmath467 ) . in @xcite , the result @xmath468 of @xcite was extended to all @xmath465 $ ] showing @xmath469 in both @xcite and @xcite , controlling the variance crucially depends on focusing on an appropriate subset of solutions ( akin to our nae - assignments , but less heavy - handed ) . in @xcite , achlioptas and naor applied the naive second moment method to the canonical symmetric constraint satisfaction problem , to the number of @xmath0-colorings of a random graph .
bearing out our belief that the naive approach should work for symmetric problems they obtained asymptotically tight bounds for the @xmath0-colorability threshold .
the difficulty there is that the overlap parameter " is a @xmath470 matrix rather than a single real @xmath471 $ ] . since @xmath472 , this makes the asymptotic analysis dramatically harder and much closer to the realm of statistical mechanics calculations . 1 .
does the second moment method give tight lower bounds on the threshold of all constraint satisfaction problem with a permutation symmetry ?
does it perform well for problems that are symmetric on average " ?
for example , does it perform well for _ regular _ random @xmath0-sat where every literal appears an equal number of times ?
3 . what rigorous connections can be made between the success of the second moment method and the notion of `` replica symmetry '' in statistical physics ? 4 . is there a polynomial - time algorithm that succeeds with uniformly positive probability close to the threshold , or at least for @xmath473 where @xmath474
? we are grateful to paul beame , ehud friedgut , michael molloy , assaf naor , yuval peres , alistair sinclair , and chris umans for reading earlier versions and making many helpful suggestions , and to remi monasson for discussions on the replica method .
we would like to thank henry cohn for bringing @xcite to our attention .
is funded partly by the national science foundation under grant phy-0200909 , and thanks tracy conrad for her support .
a. z. broder , a. m. frieze , and e. upfal . on the satisfiability and maximum satisfiability of random @xmath476-cnf formulas .
in _ proc .
4th annual symposium on discrete algorithms _ , pages 322330 , 1993 .
p. flajolet , d.e .
knuth , and b. pittel .
the first cycles in an evolving graph .
, 1 - 3 ( 1989 ) , 167215 . j. franco and m. paull .
probabilistic analysis of the davis
putnam procedure for solving the satisfiability problem . , 5(1):7787 , 1983 .
a. frieze and n. c. wormald .
random @xmath0-sat : a tight threshold for moderately growing @xmath0 . in _ proceedings of the 5th int
. symp . on theory and applications of satisfiability testing _ , pages 16 , 2002 . | many np - complete constraint satisfaction problems appear to undergo a `` phase transition '' from solubility to insolubility when the constraint density passes through a critical threshold . in all such cases
it is easy to derive upper bounds on the location of the threshold by showing that above a certain density the first moment ( expectation ) of the number of solutions tends to zero .
we show that in the case of certain symmetric constraints , considering the second moment of the number of solutions yields nearly matching lower bounds for the location of the threshold .
specifically , we prove that the threshold for both random hypergraph 2-colorability ( property b ) and random not - all - equal @xmath0-sat is @xmath1 . as a corollary
, we establish that the threshold for random @xmath0-sat is of order @xmath2 , resolving a long - standing open problem . |
in order to look for physics beyond the standard model ( sm ) , low energy experiments can play an important role . various flavor changing neutral current ( fcnc ) processes in @xmath0 and @xmath1 meson decays and the muon anomalous magnetic moment could receive large contributions from loop diagrams of new particles .
on - going experiments at kek and slac @xmath0-factories and the muon storage ring at bnl as well as future plans of @xmath0 and @xmath1 meson decay experiments therefore may be able to reveal physics beyond the energy scale which is currently accessible from direct particle search experiments .
there have been extensive studies on fcnc effects in the supergravity model . in a supersymmetry ( susy ) model , the susy particles can give large contributions to various fcnc processes such as @xmath14@xcite , @xmath2@xmath3 and @xmath4@xmath5 mixings , @xcite and @xmath15 @xcite .
these contributions depend on the flavor mixing in the squark mass matrix which is _ a priori _ independent of the flavor mixing in the quark sector .
in fact , fcnc processes can put strong constraints on these mixing parameters because arbitrary mixing parameters tend to give too large effects on , for instance , the @xmath4@xmath5 mixing amplitude if the susy particles exist below 1 tev . in the minimal supergravity model ,
these dangerous fcncs are eliminated by the assumption that all squarks have universal susy breaking mass terms through the flavor - blind coupling of gravity interaction . in the previous publication @xcite
we have investigated the @xmath2@xmath3mixing , the cp violating parameter of @xmath4@xmath5 mixing , @xmath6 , @xmath16 , @xmath17 and @xmath18 in the model based on supergravity taking @xmath14 as one of phenomenological constraints .
susy effects can increase the @xmath2@xmath3 mixing amplitude and @xmath6 by up to 20% in the minimal supergravity model .
if we relax the strict universality of all scalar susy breaking masses at the gut scale so that we take the soft susy breaking scalar masses for the higgs fields different from that for squarks and sleptons , the maximal deviation becomes about 40% .
the branching ratios for @xmath19 and @xmath20 can be suppressed by up to 10% compared to the sm predictions for the relaxed initial conditions while the deviation becomes only a few % in the minimal supergravity case . in this paper , we perform quantitative study of fcnc processes in the supergravity model taking account of updated constraints on susy parameter space . among recent improvements on susy searches ,
the susy higgs boson search in the lep ii experiment turns out to give the most significant effect on constraining the susy parameter space .
we show that possible deviation of the @xmath2@xmath3 mixing amplitude from the sm prediction becomes significantly reduced by the new constraint imposed by the higgs boson search , especially for a low @xmath7 region . taking into account the recent result of the higgs boson search ,
the maximal deviation of the @xmath2@xmath3 mixing from the sm prediction is 12% in the minimal case and 25% in the nonminimal case for @xmath21 . if lep ii does not find the lightest cp - even higgs boson below 105 gev , the deviation becomes 11% and 16% , respectively .
we also consider the muon anomalous magnetic moment in the supergravity model .
there have been many works on the muon anomalous magnetic moment in the mssm @xcite and in the supergravity model @xcite .
it is known that this process gives a sizable deviation from the sm for large @xmath7 . at present , a possible new physics contribution to @xmath22 is estimated as @xmath23 @xcite . for the present higgs mass bound , we find that the susy contribution to @xmath22 can be @xmath24 @xmath25 for @xmath26 and therefore the muon anomalous magnetic moment becomes a useful constraint on the susy parameters when the on - going experiment improves the sensitivity by a factor of 1020 .
the rest of this paper is organized as follows . in sec .
[ sec : model ] , the supergravity model is introduced and the phenomenological constraints including the higgs mass bound are described . in sec . [
sec : num ] , the numerical results on the fcnc processes are presented .
the numerical results on the muon anomalous magnetic moment are shown in sec .
[ sec : g2mu ] .
[ sec : conc ] is devoted for the conclusion .
in this section we briefly describe the minimal supersymmetric standard model ( mssm ) lagrangian based on the supergravity model .
details are given in refs .
@xcite .
the field contents of mssm consist of su(3 ) , su(2 ) and u(1 ) gauge supermultiplets @xmath27 , @xmath28 and @xmath29 , respectively , the higgs supermultiplets @xmath30 and @xmath31 , and chiral supermultiplets corresponding to quarks / squarks and leptons / sleptons .
the su(3)@xmath32su(2)@xmath32u(1 ) quantum numbers are given by @xmath33 where @xmath34 is the generation index .
requiring the @xmath35-parity conservation , the mssm superpotential is given by @xmath36 the general form of the soft susy breaking terms are given by @xmath37 where @xmath38 , @xmath39 , @xmath40 , @xmath41 , @xmath42 , @xmath43 and @xmath44 are scalar components of chiral superfields @xmath45 , @xmath46 , @xmath47 , @xmath48 , @xmath49 , @xmath50 and @xmath51 , respectively , and @xmath52 , @xmath53 and @xmath54 are @xmath55 , @xmath56 and @xmath57 gauge fermions , respectively . in the supergravity model , we assume that the soft susy breaking terms have a simple structure at the planck or gut scale . here
we assume the following initial conditions for renormalization group equations ( rges ) for the soft susy breaking terms at the gut scale : ( m_q^2)^i _ j & = & ( m_e^2)^i _ j = m_0 ^ 2^i _ j , + ( m_d^2)_i^ j & = & ( m_u^2)_i^ j = ( m_l^2)_i^ j = m_0 ^ 2_i^ j , + _ 1 ^ 2 & = & _ 2 ^ 2 = _ 0 ^ 2 , + a_d^ij & = & f_d^ij a_x m_0 , a_l^ij = f_l^ij a_x m_0 , a_u^ij = f_u^ij a_x m_0 , + m_1 & = & m_2 = m_3 = m_gx .
[ softgut ] in the minimal supergravity model the soft breaking parameters @xmath58 and @xmath59 are assumed to be equal whereas in the nonminimal case we treat the two as independent parameters .
we assume that the soft susy breaking terms @xmath58 , @xmath59 , @xmath60 , @xmath61 and the @xmath62 parameter are all real so that we do not consider the constraint from electron and neutron electric dipole moments ( edms ) . with the above initial conditions we can solve the one - loop rges for the susy breaking parameters and determine these parameters at the electroweak scale .
we also require that the electroweak symmetry breaking occurs properly to give the correct @xmath63 boson mass . as independent parameters , we take @xmath7 , @xmath58 , @xmath61 , @xmath60 and @xmath64 ( and @xmath59 for the nonminimal case ) .
once we fix a set of these parameters we can calculate the masses and mixings of all susy particles .
we can then evaluate various amplitudes for fcnc processes such as @xmath14 , the @xmath2@xmath3 mixing amplitude , @xmath6 , @xmath17 and @xmath18 . in order to determine the allowed region of the susy parameter space , we require the following phenomenological constraints 1 . [ item : bsg ] @xmath9 constraint from cleo , _
i.e. _ , @xmath65 @xcite .
[ item : lightsp ] the chargino mass is larger than 97 gev , and all other charged susy particle masses should be larger than 90 gev @xcite .
3 . all sneutrino masses are larger than 41 gev @xcite .
4 . the gluino and squark mass bounds from tevatron experiments @xcite . 5 .
the lightest susy particle is neutral .
[ item : ccb ] the condition for not having a charge or color symmetry breaking minimum @xcite . 7 .
[ item : higgs ] susy higgs boson search at lep ii @xcite .
the detail of these conditions [ item : bsg][item : ccb ] is discussed in refs .
the difference between refs .
@xcite and the present analysis is that the experimental bounds in [ item : bsg ] and [ item : lightsp ] are updated and the complete next - to - leading order qcd correction formula is used for the calculation of the susy contributions to the @xmath9 amplitude @xcite .
previously in refs .
@xcite the condition [ item : higgs ] did not give any impact to constrain the susy parameter space .
due to the recent improvement of the higgs boson search at lep ii @xcite , a sizable parameter space starts to be excluded .
this will be discussed in the next subsection .
it is well known that there is a strict upper bound on the lightest cp - even higgs boson mass in the mssm @xcite . because the higgs search in the lep ii experiment has reached to the sensitivity of 100 gev for the sm higgs boson and about 90 gev for the case of the lightest cp - even higgs boson in the mssm , a meaningful amount of the susy parameter space
is already excluded @xcite .
the higgs sector in the mssm consists of five physical mass eigenstates : two cp - even higgs boson @xmath66 and @xmath67 , one cp - odd higgs boson @xmath68 and a pair of charged higgs boson @xmath69 .
there are two angles to specify the mixing in the higgs sector , namely , a ratio of two vacuum expectation values @xmath70 and the mixing angle @xmath71 for the two cp - even higgs bosons defined as @xmath72 where we have only kept the cp - even components of two higgs fields . in the mssm ,
these masses and mixings are determined by two input parameters , for instance @xmath73 and @xmath74 , as well as stop and sbottom masses and mixing parameters through one - loop corrections to the higgs potential formula . in the supergravity model these masses and mixings at the electroweak scale are calculated by solving the rges from the initial conditions at the gut scale .
the constraints on the susy parameter space from the lep experiments are expressed as an allowed region of two dimensional parameters , for instance @xmath75 , with different assumptions of the remaining parameters ( stop mass , _ etc . _ ) . in order to simplify the treatment in the numerical analysis we use the following procedure .
we first take the experimental constraint in the space of @xmath73 and @xmath74 from ref .
we require that the set of @xmath73 and @xmath74 calculated in the supergravity model is within the allowed region of this parameter space .
strictly speaking , the allowed region slightly depends on the stop mass and mixing parameters through the angles @xmath71 and @xmath76 , but this dependence is not significant .
in addition , except for a small parameter region in the case of a large @xmath7 , the cp - odd scalar mass becomes much larger than @xmath77 so that the property of the lightest cp - even higgs boson becomes very close to that of the sm higgs boson .
in such a case the susy higgs boson search essentially gives the same lower bound of the lightest cp - even higgs boson mass as that of the sm higgs boson , and therefore almost independent of other susy parameters . in the numerical calculation we use the analytic formula for the cp - even higgs mass matrix given in ref .
this formula takes into account two - loop leading - log corrections and reproduces the next - to - leading - log results within a few gev for @xmath78 tev . as the present experimental constraint we require that the set of @xmath73 and @xmath74 is within the allowed region shown in fig .
[ fig : higgs ] ( case ( i ) ) .
later we also show how various results depend on a possible future improvement of the lep ii higgs boson search .
we then simply require @xmath79 gev to get a rough idea for the case that the lep ii experiment will not discover any light higgs boson ( case ( ii ) ) .
in this section , the results of numerical calculations for various fcnc processes in @xmath0 and @xmath1 decays are presented .
we consider here 1 .
the @xmath2@xmath3 mixing amplitude ; 2 .
@xmath6 ; 3 .
@xmath17 ; 4 .
@xmath18 .
detail descriptions on the calculation of these quantities in the supergravity model are given in ref .
@xcite . besides the improvement discussed in sec .
[ sec : model ] , we have improved the calculation of the @xmath2@xmath3 and @xmath4@xmath5 mixings by including the next - to - leading - order qcd correction to the contribution due to the charged - higgs top loop diagram@xcite .
we first present the magnitude of @xmath80 in the supergravity model . in fig .
[ fig : xd - mhl ] we show @xmath81 as a function of the lightest higgs boson mass for three values of @xmath7 , _
@xmath82 , where @xmath83 denotes the sm value . as input parameters
we take @xmath84 gev , @xmath85 gev , @xmath86 . for the ckm matrix elements ,
we adopt the standard phase convention of the particle data group @xcite , taking @xmath87 , @xmath88 , @xmath89 .
we fix the cp violating phase in the ckm matrix @xmath90 in the following analysis . as we will see later , the ratio @xmath81 is almost independent of @xmath91 .
the susy breaking parameters at the gut scale are scanned within the following range : @xmath92 tev , @xmath93 tev , @xmath94 tev , @xmath95 .
this figure shows the possible values of @xmath81 and @xmath74 for parameter sets of the minimal supergravity model as well as for those of the nonminimal case discussed in sec .
[ sec : model ] . campared to the result in @xcite , we see that the possible deviation from the sm is reduced to @xmath96% from @xmath97% for the nonminimal case if we require that the lightest cp - even higgs mass should be larger than 95 gev for @xmath98 .
if the mass bound is raised to 105 gev no allowed parameter space remains for @xmath98 . for @xmath99 ,
the lightest higgs boson mass is shifted to a larger value so that there is some allowed space for @xmath100 gev . for @xmath26 ,
a separate allowed region appear in the region of @xmath101125 gev where @xmath81 is 1.151.30 .
it is instructive to see the correlation between @xmath81 and @xmath14 . in fig .
[ fig : xd - bsg ] we show @xmath81 as a function of @xmath14 for @xmath82 . in these figures ,
the constraint from the present susy higgs boson search is applied as described in sec .
[ sec : higgs ] . for @xmath102 ,
most of the parameter region is already excluded by the higgs boson search .
the deviations of both @xmath14 and @xmath80 from the sm predictions are small in the allowed parameter region . compared with the fig.8(c ) in ref .
@xcite , we can see that the more parameter space is excluded in the case that @xmath14 is larger than the sm prediction .
this is because both @xmath14 and the mass of the lightest cp - even higgs boson have a correlation with @xmath64 .
namely , @xmath14 is enhanced for @xmath103 . on the other hannd
the lightest higgs boson becomes lighter for @xmath103 since it depends on the @xmath64 through the stop left - right mixing parameter in the radiative correction part of the higgs mass formula . for @xmath26 , there are two branches within the allowed region of @xmath14 for the nonminimal cese .
the upper dots correspond to the allowed region where @xmath101125 gev in fig .
[ fig : xd - mhl](c ) . in these points
the sign of the @xmath9 amplitude becomes opposite to that of the @xmath9 amplitude in the sm because the susy contributions are very large and opposite in sign .
it is known that this parameter space corresponds to the case where the @xmath104 branching ratio is enhanced by 50100% @xcite . in fig .
[ fig : xd - tanb](a ) the allowed ranges of @xmath81 are shown for several values of @xmath7 with the present constraint of the susy higgs boson search and fig .
[ fig : xd - tanb](b ) corresponds to the case where @xmath100 gev . for @xmath105
there appears a separate region where the sign of the @xmath9 amplitude is opposite to that of the sm amplitude .
two cases are distinguished according to the sign of the @xmath9 amplitude for @xmath106 .
we can see that the present constraint allows the deviation of the sm for the @xmath2@xmath3 mixing up to 12% for the minimal case and 30% for the nonminimal case .
if the higgs boson bound is raised to 105 gev , the allowed deviation from the sm becomes less than 15% except for the small parameter space where the sign of the @xmath9 amplitude becomes opposite to the sm case .
next we show the branching ratio of @xmath19 normalized to the sm prediction .
we have calculated @xmath107 in the minimal and nonminimal cases .
the allowed range of this quantity is shown for the present higgs bound ( fig .
[ fig : klpnn - tanb](a ) ) and for @xmath108 gev ( fig .
[ fig : klpnn - tanb](b ) ) .
we can see that @xmath17 is suppressed by up to 5% and 15% in the minimal and nonminimal cases , respectively .
when we restrict @xmath100 gev , the maximal deviation is reduced to 7 % for the nonminimal case .
we study various correlations among @xmath80 , @xmath6 , @xmath17 and @xmath18 . as noted in ref .
@xcite , @xmath81 and @xmath109 have a linear relation .
when these quantities are normalized by the corresponding quantities in the sm , these relations are essentially independent of @xmath91 in the ckm matrix because the factors of ckm matrix elements are approximately equal for @xmath80 and @xmath83 . for @xmath6 and @xmath110 ,
there is a charm loop contribution which breaks the proportionality , but this contribution turns out to be small . in fig .
[ fig : xd - ek ] the correlation between @xmath81 and @xmath109 is shown for @xmath8 and the nonminimal case . the corresponding figure for the minimal case
is obtained by taking a part of these lines according to the maximal value of @xmath81 . the correlation between @xmath107 and @xmath111
is shown in fig .
[ fig : kppnn - klpnn ] for @xmath8 and the nonminimal case .
we can see that the situation is quite similar to fig .
[ fig : xd - ek ] .
the slight dependence on @xmath91 is the result of the charm loop contribution to the @xmath112 .
we also show the correlation between @xmath81 and @xmath107 in fig .
[ fig : klpnn - xd ] for @xmath8 and 10 with the present higgs search limit .
we can see that in the parameter space where @xmath80 is the most enhanced the suppression of @xmath17 become maximal .
the correlations are useful to distinguish possible new physics effects from consistency check of the unitarity triangle . for instance , suppose that the time - dependent cp asymmetry in the @xmath113 decay mode is well established . then by combining one more observable quantity , the parameter set @xmath114 of the ckm matrix in the wolfenstein parametrization is determined within the sm . if the susy effects are relevant , @xmath115 determined from @xmath80 or @xmath6 can be different from that determined from @xmath17 or @xmath18 because the formers are enhanced and the latters are suppressed in the supergravity model . on the other hand , observables such as @xmath116 , @xmath117 , and
cp asymmetries in @xmath0 decays are essentially independent of the susy loop contributions .
in this section we present the numerical result of susy loop effects on the muon anomalous magnetic moment in the supergravity model .
the present experimental value of @xmath118 is @xcite @xmath119 according to ref .
@xcite the sm prediction is @xmath120 where the error in the sm value is dominated by the hadronic contribution of the vacuum polarization diagram @xcite . combining the above two values
we can derive a possible new physics contribution to @xmath22 as @xmath121 the current bnl experiment is aiming to improve @xmath22 by a factor of 20 and the first result is reported as @xmath122 @xcite .
in addition the hadronic contribution to the vacuum polarization may be better understood if the @xmath123 total cross section is well measured experimentally @xcite . as we see later , although the present constraint from the muon anomalous magnetic moment is not strong enough the situation will soon change after the improvement of the measurement .
we calculated the susy contribution to @xmath22 ( @xmath124 ) from the loop diagrams with sneutrino and chargino and with charged slepton and neutralino .
detailed formula are found , for example , in @xcite .
we require the radiative electroweak symmetry breaking condition and the various phenomenological constraints discussed in sec .
[ sec : model ] .
we show @xmath124 in the minimal supergravity for @xmath8 as a function of the lighter chargino ( @xmath125 ) mass , the left - handed smuon ( @xmath126 ) mass and @xmath14 in fig .
[ fig : g-2mu.03 ] . also the same plots for @xmath127
are shown in fig .
[ fig : g-2mu.10 ] . in these figures
the present bound from the higgs boson search is applied .
we can see that @xmath124 can be large only when both @xmath125 and @xmath126 are relatively light .
for example , @xmath125 and @xmath126 are lighter than 150 gev and 200 gev , respectively , in order for @xmath128 for @xmath26 .
as is well known , the magnitude of @xmath124 becomes large for large @xmath7 .
the enhancement of @xmath124 for large @xmath7 comes from the fact that @xmath124 is dominated with the sneutrino - chargino loop diagram , which contains a contribution proportional to @xmath129 . as a result ,
susy contributions to both @xmath9 amplitude and @xmath124 are correlated with @xmath64 .
we can see that @xmath124 becomes positive ( negative ) according to the suppression ( enhancement ) of @xmath14 .
this correlation was pointed out in ref .
@xcite .
the predicted range of @xmath124 are shown for several values of @xmath7 with the present constraint on the higgs search and with @xmath100 gev for the minimal and the nonminimal cases in fig .
[ fig : g-2mu - tanb ] . as in fig .
[ fig : xd - tanb ] two cases according to the sign of the @xmath9 amplitude are shown separately for @xmath130 .
this figure shows that the muon anomalous magnetic moment is indeed expected to become very powerful to constrain the susy parameter space in near future .
we see that even for @xmath131 the deviation is quite sizable considering future improvements on the @xmath22 measurement .
we updated the numerical analysis of fcnc processes in @xmath0 and @xmath1 decays and the muon anomalous magnetic moment in the supergravity model . taking account of the recent progress in the higgs boson search , we show that a small @xmath7 region is almost excluded for @xmath132 .
the maximal deviation from the sm value in the @xmath2@xmath3 mixing is 12% for the minimal supergravity case and 30% for the nonminimal case .
if the higgs mass bound is raised to 105 gev the deviations become less than 11% and 16% , respectively , except for a small parameter space where the @xmath9 decay amplitude is opposite in sign to the sm amplitude . for @xmath15 , we show that the deviation is less than 5% for the minimal case and less than 14% for the nonminimal case under the present higgs mass bound .
we also calculate the susy contribution to the muon anomalous magnetic moment .
@xmath124 and @xmath14 show a strong correlation and @xmath124becomes very large for a large @xmath7 region .
we find that the susy contribution @xmath124 can be @xmath24
@xmath25 for @xmath26 for the minimal supergravity case . along with the @xmath14 constraint
, @xmath124 will soon become a very important constraint on the parameter space in the supergravity model .
the authors would like to thank s. yamashita for useful discussions on higgs boson search experiments at lep .
the work of y.o . was supported in part by the grant - in - aid of the ministry of education , science , sports and culture , government of japan ( no.09640381),priority area `` supersymmetry and unified theory of elementary particles '' ( no .
707 ) , and `` physics of cp violation '' ( no.09246105 ) .
99 f. gabbiani and a. masiero , _ nucl .
phys . _ * b322 * , 235 ( 1989 ) ; + i.i .
bigi and f. gabbiani , _ nucl .
phys . _ * b352 * , 309 ( 1991 ) .
s. bertolini , f. borzumati , a. masiero and g. ridolfi , _ nucl .
phys . _ * b353 * , 591 ( 1991 ) .
n. oshimo , _ nucl .
* b404 * , 20 ( 1993 ) ; + j. hewett , _ phys .
* 70 * , 1045 ( 1993 ) ; + v. barger , m. berger and r.j.n .
phillips , _ phys .
lett . _ * 70 * , 1368 ( 1993 ) ; + r. barbieri and g.f .
giudice , _ phys .
* b * _ * 309 * , 86 ( 1993 ) ; + j.l .
lopez , d.v .
nanopoulos and g.t .
park , _ phys .
rev . * d * _ * 48 * , 974 ( 1993 ) ; + j.l .
lopez , d.v .
nanopoulos , g.t .
park and a. zichichi , _ phys . rev . * d * _ * 49 * , 355 ( 1994 ) ; + y. okada , _ phys . lett . * b * _ * 315 * , 119 ( 1993 ) ; + r. garisto and j.n .
ng , _ phys . lett .
* b * _ * 315 * , 372 ( 1993 ) ; + m.a .
diaz , _ phys .
* b * _ * 322 * , 207 ( 1994 ) ; + f.m .
borzumati , _ z. phys . * c * _ * 63 * , 291 ( 1994 ) ; + s. bertolini and f. vissani , _ z. phys . * c * _ * 67 * , 513 ( 1995 ) ; + j. wu , p.
nath and r. arnowitt , _ phys .
rev . * d * _ * 51 * , 1371 ( 1995 ) ; + p.
nath and r. arnowitt , _ phys .
lett . * b * _ * 336 * , 395 ( 1994 ) ; _ phys . rev .
lett . _ * 74 * , 4592 ( 1995 ) ; + f.m .
borzumati , m. drees and m.m .
nojiri , _ phys .
d * _ * 51 * , 341 ( 1995 ) ; + j.l .
lopez , d.v .
nanopoulos , x. wang and a. zichichi , _ phys .
rev . * d * _ * 51 * , 147 ( 1995 ) ; + g. kane , c. kolda , l. roszkowski and d. wells , _ phys . rev . *
d * _ * 49 * , 6173 ( 1994 ) ; + m. carena , m. olechowski , s. pokorski and c.e.m .
wagner , _ nucl .
_ * b426 * , 269 ( 1994 ) ; + c. kolda , l. roszkowski , d. wells and g. kane , _ phys .
* d * _ * 50 * , 3498 ( 1994 ) ; + v. barge , m.s .
berger , p. ohmann and r.j.n .
phillips , _ phys .
* d * _ * 51 * , 2438 ( 1995 ) ; + b. de carlos and j.a .
casas , _ phys .
* b * _ * 349 * , 300 ( 1995 ) ; erratum , _ ibid .
_ * b351 * , 604 ( 1995 ) ; + t. goto and y. okada , _ prog .
_ * 94 * , 407 ( 1995 ) ; _ prog .
suppl . _ * 123 * , 213 ( 1996 ) ; + g.t . park , _ mod .
a * _ * 11 * , 1187 ( 1996 ) ; + h. baer and m. brhlik , _ phys . rev . * d * _ * 55 * , 3201 ( 1997 ) ; + h. baer , m. brhlik d. castano and x. tata , _ phys . rev . * d * _ * 58 * , 015007 ( 1998 ) .
l.j . hall , v.a .
kostelecky and s. raby , _ nucl .
phys . _ * b267 * , 415 ( 1986 ) ; + t. kurimoto , _ phys . rev .
* d * _ * 39 * , 3447 ( 1989 ) ; _ mod .
a * _ * 10 * , 1577 ( 1995 ) ; + j.s .
hagelin , s. kelley and t. tanaka , _ mod .
a * _ * 8 * , 2737 ( 1993 ) ; _ nucl .
* b415 * , 293 ( 1994 ) ; + g.c .
branco and g.c .
cho , y. kizukuri and n. oshimo , _ phys .
b * _ * 337 * , 316 ( 1994 ) ; _ nucl .
* b449 * , 483 ( 1995 ) ; + m.p .
worah , _ phys .
d * _ * 54 * , 2198 ( 1996 ) ; + g.t . park and s.k .
kang , _ phys .
* d * _ * 54 * , 4687 ( 1996 ) ; + m. ciuchini , e. franco , g. martinelli and a. masiero , _ phys . rev . lett .
_ * 79 * , 978 ( 1997 ) ; + j.a .
bagger , k.t .
matchev and r.j .
zhang , _ phys .
* b * _ * 412 * , 77 ( 1997 ) ; + r. barbieri and a. strumia , _ nucl .
_ * b508 * , 3 ( 1997 ) ; + t. nihei , _ prog .
_ * 98 * , 1157 ( 1997 ) ; + m. ciuchini , _ et al .
_ , _ jhep _ * 9810 * , 008 ( 1998 ) ; + s. baek , j .- h . jang , p. ko , j.h .
park , hep - ph/9907572 .
t. goto , t. nihei and y. okada , _ phys .
rev . * d * _ * 53 * , 5233 ( 1996 ) ; _ phys . rev .
* d * _ * 54 * , 5904(e ) ( 1996 ) .
s. bertolini and a. masiero , _ phys .
* b * _ * 174 * , 343 ( 1986 ) ; + b. mukhopadhyaya and a. raychaudhuri , _ phys .
* b * _ * 189 * , 203 ( 1987 ) ; + i.i .
bigi and f. gabbiani , _ nucl .
* b367 * , 3 ( 1991 ) ; + g. couture and h.knig , _ z. phys .
* c * _ * 69 * , 174 ( 1995 ) ; + y. nir and m.p .
worah , _ phys .
* b * _ * 423 * , 319 ( 1998 ) ; + a.j .
buras , a. romanino and l. silvestrini , _ nucl .
* b * _ * 520 * , 3 ( 1998 ) ; + g.c .
cho , _ eur .
j. _ * c5 * , 525 ( 1998 ) ; + g. colangelo , g. isidori , _ jhep _ * 9809 * , 009 ( 1998 ) ; + a.j .
buras , g. colangelo , g. isidori , a. romanino , l. silvestrini , hep - ph/9908371 . t. goto , y. okada and y. shimizu , _ phys . rev .
* d * _ * 58 * , 094006 ( 1998 ) .
p. fayet , in unification of the fundamental particles interactions , edited by s. ferrara , j. ellis and p. van nierwenhuizen ( plenum , new york , 1980 ) p. 587 ; + j.a .
grifols and a. mendez , _ phys .
rev . * d * _ * 26 * , 1809 ( 1982 ) ; + j. ellis , j.s . hagelin and d.v .
nanopoulos , _ phys .
* b * _ * 116 * , 283 ( 1982 ) ; + r. barbieri and l. maiani , _ phys .
* b * _ * 117 * , 203 ( 1982 ) ; + m. carena , g.f .
giudice , and c.e.m .
wagner , _ phys .
* b * _ * 390 * , 234 ( 1997 ) ; + t. ibrahim and p. nath , hep - ph/9908443 .
t. moroi , _ phys .
* d * _ * 53 * , 6565 ( 1996 ) ; erratum - ibid . * d56 * , 4424 ( 1997 ) . d.a .
kosower , l.m . krauss and n. sakai ,
lett . * b * _ * 133 * , 305 ( 1983 ) ; + t.c . yuan , r. arnowitt , a.h .
chamseddine and p. nath , _
z. phys . * c * _ * 26 * , 407 ( 1984 ) ; + j.c .
romao , a. barroso , m.c .
bento and g.c .
branco , _ nucl .
* b * _ * 250 * , 295 ( 1985 ) ; + j.l .
lopez , d.v .
nanopoulos and x. wang , _ phys .
rev . * d * _ * 49 * , 366 ( 1994 ) ; + t. ibrahim and p. nath , hep - ph/9907555 .
u. chattopadhyay and p. nath _ phys .
* d * _ * 53 * , 1648 ( 1996 ) . a. czarnecki and w.j .
marciano , he - ph/9810512 .
t. goto , y. okada , y. shimizu and m. tanaka , _ phys .
rev . * d * _ * 55 * , 4273 ( 1997 ) .
cleo collaboration , hep - ex/9908022 .
opal collaboration , opal physics note pn413 ( 1999 ) .
v. ruhlmann - kleider , talk given in xix international symposium on lepton and photon interactions at high energies , stanford university , august 9 - 14 , 1999 .
et al . _ , ( particle data group ) , _ eur .
j. _ * c3 * , 1 ( 1998 ) .
cdf collaboration , f. abe _ et al .
_ , _ phys .
d * _ * 56 * , 1357 ( 1997 ) ; + d0 collaboration , s. abachi _ et al .
_ , _ phys .
_ * 75 * , 618 ( 1995 ) .
derendinger and c.a .
savoy , _ nucl .
phys . _ * b237 * , 307 ( 1984 ) .
aleph collaboration , hep - ex/9908016 ; + opal collaboration , opal physics note pn414 ( 1999 ) . c. bobeth , m. misiak and j. urban , tum - hep-321 - 98 , hep - ph/9904413 . y. okada , m. yamaguchi and t. yanagida , _ prog . theor . phys . _ * 85 * , 1 ( 1991 ) ; _ phys . lett .
* b * _ * 262 * , 54 ( 1991 ) ; + j. ellis , g. ridolfi and f. zwirner , _ phys . lett . * b * _ * 257 * , 83 ( 1991);_phys .
* b * _ * 262 * , 477 ( 1991 ) ; + h.e . haber and r. hempfling , _ phys . rev .
lett . _ * 66 * , 1815 ( 1991 ) .
m. carena , j.r .
espinosa , m. quirs and c.e.m .
wagner , _ phys .
* b * _ * 355 * , 209 ( 1995 ) . j. urban , f.krauss , u. jentschura , g. soff , _ nucl .
* b * _ * 523 * , 40 ( 1998 ) . r.m .
carey _ et al .
. rev .
* 82 * , 1632 ( 1999 ) .
e. de rafael , _ phys .
* b * _ * 322 * , 239 ( 1994 ) ; + s. eidelman and f. jegerlehner , _ z. phys
. * c * _ * 67 * , 585 ( 1995 ) ; + j. bijnens , e. pallante , and j. prades , _ nucl . phys . * b * _ * 474 * , 379 ( 1996 ) ; + b. krause , _ phys .
lett . * b * _ * 390 * , 392 ( 1997 ) ; + m. hayakawa and t. kinoshita , _ phys . rev . *
d * _ * 57 * , 465 ( 1998 ) ; + r. alemany , m. davier , and a. hocker , _ eur .
j. _ * c2 * , 123 ( 1998 ) .
m. davier and a. hocker , _ phys .
* b * _ * 435 * , 427 ( 1998 ) .
* fig . * allowed region in @xmath73@xmath74 space used in our numerical calculation .
[ fig : higgs ] @xmath81 as a function of the lightest higgs boson mass for ( a ) @xmath102 , ( b ) @xmath8 and ( c ) @xmath127 .
each dot represents the value in the full parameter space and each square shows the value for the minimal case . [
fig : xd - mhl ] @xmath81 as a function of @xmath14 for ( a ) @xmath102 , ( b ) @xmath8 and ( c ) @xmath127 .
each dot represents the value in the full parameter space and each square shows the value for the minimal case .
the vertical dotted lines show the upper and lower bounds on @xmath14 given by cleo .
the constraint from the present higgs boson search ( case ( i ) ) is imposed .
[ fig : xd - bsg ] allowed ranges of @xmath81 for several values of @xmath7 with ( a ) the present constraint of the higgs boson mass ( case ( i ) ) and ( b ) @xmath100 gev ( case ( ii ) ) .
two lines are shown according to the sign of the @xmath9 amplitude for @xmath133 .
the left ( right ) lines correspond the case where the sign of the @xmath9 amplitude is same ( opposite ) as that in the sm .
[ fig : xd - tanb ] allowed ranges of @xmath107 for several values of @xmath7 with ( a ) the present constraint of the higgs boson mass ( case ( i ) ) and ( b ) @xmath100 gev ( case ( ii ) ) .
the allowed ranges for the parameter regions where the sign of the @xmath9 decay amplitude is opposite to that of the sm amplitude are separately plotted .
the meaning of two lines for @xmath133 is as the same as in the case of fig .
[ fig : xd - tanb ] .
[ fig : klpnn - tanb ] correlation between @xmath81 and @xmath109 for @xmath134 , @xmath135 and @xmath136 in the nonminimal case with @xmath8 with the present constraint of the higgs boson mass ( case ( i ) ) .
[ fig : xd - ek ] correlation between @xmath107 and @xmath111 for @xmath134 , @xmath135 and @xmath136 in the nonminimal case with @xmath8 with the present constraint of the higgs boson mass ( case ( i ) ) .
[ fig : kppnn - klpnn ] correlation between @xmath107 and @xmath81 for ( a ) @xmath137 and ( b ) @xmath138 with the present constraint of the higgs boson mass ( case ( i ) ) .
[ fig : klpnn - xd ] @xmath124 in the minimal supergravity case for @xmath8 ( a ) as a function of the lighter chargino mass , ( b ) as a function of the left - handed scalar muon mass , and ( c ) as a function of @xmath14 .
the constraint from the present higgs boson search ( case ( i ) ) is imposed .
[ fig : g-2mu.03 ] same as fig .
[ fig : g-2mu.03 ] for @xmath127 .
[ fig : g-2mu.10 ] allowed ranges of @xmath124 for several values of @xmath7 with ( a ) the present constraint of the higgs boson mass ( case ( i ) ) and ( b ) @xmath100 gev ( case ( ii ) ) .
the allowed ranges for the parameter regions where the sign of the @xmath9 decay amplitude is opposite to that of the sm amplitude are separately plotted .
[ fig : g-2mu - tanb ] | flavor changing neutral current processes in @xmath0 and @xmath1 decays in the supergravity model are revisited taking into account the recent progress of higgs boson search experiments at lep .
possible deviations of @xmath2@xmath3 mixing , the cp violating @xmath4@xmath5 mixing parameter @xmath6 from the standard model predictions are reduced significantly in a small @xmath7 region due to the constraints on the susy parameter space imposed by the higgs boson search . with the present bound on the higgs boson mass ,
the magnitude of the @xmath2@xmath3 mixing amplitude can be enhanced up to 12% for @xmath8 in the minimal supergravity .
if we relax the strict universality of susy breaking scalar masses at the gut scale , the deviation can be 25% .
we also investigate susy contributions to the muon anomalous magnetic moment in the supergravity model and show that there is a correlation with the @xmath9 amplitude .
the susy contribution to @xmath10 can be ( @xmath11 @xmath12 for @xmath13 and even larger for larger @xmath7 .
the improved measurement of the muon anomalous magnetic moment in the on - going experiment at bnl will put a strong constraint on susy parameter space , especially for a large @xmath7 region .
* * [ cols= " < " , ] * flavor changing neutral current processes and the muon anomalous magnetic moment in the supergravity model * + + |
solid - state q - switched oscillators allowing nano- and sub - nanosecond pulsing find applications in a lot of areas including medicine , spectroscopy , environment monitoring , etc .
passive q - switching based on the use of both semiconductor @xcite and crystalline @xcite saturable absorbers is of particular interest due to compactness , simplicity , high damage threshold , and diode - pumping ability of an oscillator . among the active media allowing diode - pumped q - switching , ky(wo@xmath0)@xmath1 ( kyw ) and
kgd(wo@xmath0)@xmath1 ( kgw ) crystals doped by yb@xmath2 and nd@xmath2 ions are known as the materials providing an efficient stimulated raman scattering ( srs ) @xcite . as a result
, there is possible a high - efficient self - frequency shift and a simultaneous two - wavelengths pulsing ( e.g. , at 1.35 and 1.54 @xmath3 m for a nd : kgw active medium ) directly from an oscillator .
theoretical studies of a q - switched oscillator with the intra - cavity srs have been based on the well - established rate - equations approach @xcite .
the oscillator parameters providing a pulse width minimization @xcite and an output energy maximization @xcite at both fundamental and stokes wavelengths have been defined .
simultaneously , it has been found that the spatial structure of a laser field is strongly affected by the srs @xcite .
this means that the mode transformation has to be taken into account along with the temporal evolution of fields and populations inside an active medium and a saturable absorber .
such a model has been developed to a gaussian mode approximation for both fundamental and stokes fields @xcite .
this has allowed defining the optimal values of the raman gain and the ratio of pump and laser beams , which provide a single pulse operation of oscillator with the most efficient srs .
nevertheless , a transient character of q - switching can prevent from the cw mode formation and disturb substantially the spatial structure of a laser field .
this requires to take into account the spatial dynamics on a par with the temporal one . in this work we present the analysis of spatially - temporal dynamics of a passively q - switched raman - active oscillator .
the analysis is based on a rate - equations approach but without imposing a limitation on the shape of cylindrically symmetric transverse distribution of a field .
the results demonstrate that there exist both spatial extra - broadening and squeezing as well as transition between non - gaussian and gaussian spatial profiles in dependence on the oscillator parameters ( viz .
output mirror transmission at the fundamental and stokes wavelengths , initial transmission of a saturable absorber and curvature of a spherical mirror ) .
the srs contribution controlled by the optimized sets of oscillator parameters allows the regularization of the spatially - temporal structure of a field and the substantial pulse shortening .
the oscillator under consideration consists of the flat ( output ) and spherical mirrors .
the active medium ( nd@xmath2:kgw ) is placed at the resonator center , and the saturable absorber ( v@xmath2:yag ) is placed on the output mirror .
the model is based on a direct generalization of that presented in @xcite .
it is supposed that the time - dependent ( @xmath4 is the time , which is periodical with the cavity period @xmath5 : @xmath6 } $ ] ) fields @xmath7 ( indexes @xmath8 and @xmath9 correspond to the fundamental and stokes fields , respectively ) have the radially - symmetric transverse spatial distributions ( @xmath10 { 0,r } \right ] } $ ] is the radial coordinate ) . inside an active medium , the dynamics is modeled on basis of the split - step hankel s method .
that is the active medium volume is considered to be divided into 10 transverse slices with the thickness @xmath11 ( @xmath12 is the longitudinal coordinate ) , in which the dynamics within the time - domain and the domain of spatial frequencies ( @xmath13 , @xmath14 is the velocity of light , @xmath15 cm ) @xcite is evolved step - by - step : @xmath16 \hfill \\ \dot a_s \left ( { z , r , t } \right ) = \frac{{\delta z } } { 2 t_{cav}}a_s \left ( { z , r , t } \right)g_s \left| { a_f \left ( { z , r , t } \right ) } \right|^2 \hfill \\ \dot g\left ( { r , t } \right ) = - \frac{{\sigma_g \lambda_f } } { h c } g\left ( { r , t } \right)\left| { a_f \left ( { z , r , t } \right ) } \right|^2 \hfill \\ \end{gathered } \right\ } \otimes\\ \nonumber \otimes \tilde a_{f , s}\left ( { z+\delta z,\nu , t } \right ) = \tilde a_{f , s}\left ( { z,\nu , t } \right ) \exp \left ( { - ik_{f , s } \delta z + \frac{i } { 2}k_{f , s } \delta z\lambda _ { f , s}^2 \nu ^2 } \right).\end{aligned}\ ] ] we shall suppose , that the field intensities @xmath17 are normalized to @xmath18 ( @xmath19 is the planck constant ; @xmath201.35 @xmath3 m and @xmath211.54 @xmath3 m are the fundamental and stokes wavelengths , respectively ; @xmath220.76@xmath23 @xmath24 is the gain cross - section , @xmath250.8 ns ) , and the time is normalized to @xmath5 .
the gain coefficient is @xmath26 and its initial value equals to the threshold one : @xmath27 . in the last expression , the varied values @xmath28 and @xmath29
correspond , respectively , to the initial transmission of saturable absorber and the reflection of output mirror at the fundamental wavelength ; @xmath30 is the unsaturable net - loss coefficient .
@xmath315 cm is the active medium length .
the stimulated raman scattering inside an active medium is described by the coefficient @xmath32 and its dimensional value amounts to 6 cm / gw . it is assumed that there exists no anti - stokes and higher - order stokes scattering as well as that the gain is saturable by only fundamental field .
also , the initial gain distribution is assumed to be spatially homogeneous .
the field representations @xmath33 within the spatial frequency domain are obtained by the means of the fast hankel s transformation @xcite .
@xmath34 are the wave - numbers corresponding to the fundamental and stokes fields , respectively .
a similar procedure describes the dynamics inside a saturable absorber : @xmath35 \hfill \\ \dot a_s \left ( { z , r , t } \right ) = - \frac{{\delta z } } { 2 t_{cav}}a_s \left ( { z , r , t } \right)n\left ( { r , t } \right ) \hfill \\ \dot n\left ( { r , t } \right ) = - \frac{{\sigma_f \lambda_f } } { h c } n\left ( { r , t } \right)\left| { a_f \left ( { z , r , t } \right ) } \right|^2 - \frac{{\sigma_s \lambda_f } } { h c } n\left ( { r , t } \right)\left| { a_s \left ( { z , r , t } \right ) } \right|^2 + \frac{{n - n\left ( { r , t } \right ) } } { { t_r } } \hfill \\ \end{gathered } \right\ } \otimes\\ \nonumber \otimes \tilde a_{f , s}\left ( { z+\delta z,\nu , t } \right ) = \tilde a_{f , s}\left ( { z,\nu , t } \right ) \exp \left ( { - ik_{f , s } \delta z + \frac{i } { 2}k_{f , s } \delta z\lambda _ { f , s}^2 \nu ^2 } \right)\end{aligned}\ ] ] here , @xmath36 is the loss coefficient and its initial value amounts to @xmath37 ( @xmath38=0.1 cm is the saturable absorber thickness ) .
the loss is saturable by both fundamental and stokes fields ( @xmath39 , @xmath40 , when the normalization to @xmath41 is presumed ) .
excited - state absorption is taken into account for the fundamental field ( @xmath420.1 ) .
@xmath43 is the loss relaxation time ( the normalization to @xmath5 is presumed ) .
free propagations between the active medium and the absorber as well as between the active medium and the spherical mirror are considered in the frequency domain : @xmath44 where @xmath45 cm is the propagation length .
the reflection from a spherical mirror obeys : @xmath46 where @xmath47 is the variable radius of the mirror curvature .
the loss on an output mirror as well as the net unsaturable loss are taken into account by means of the following mapping : @xmath48,\ ] ] where @xmath49 is the output mirror reflection at the stokes wavelength . in the simulations ,
the time window is approximated by 100 points and the window of transverse coordinate is approximated by 2000@xmath506000 points in dependence on the @xmath51 value , which varies from 0.5 to 1.5 cm in order to exclude the boundary effects .
the solution in the time domain is evaluated using a fourth - order runge - kutta method .
the initial dimensionless fields have the gaussian spatial profile with the size , which equals to that of the fundamental cw mode at @xmath52 .
the initial dimensionless amplitudes @xmath53 and @xmath54 amount to 10@xmath55 and 10@xmath56 , respectively .
the simulations based on the model described in the previous section demonstrate that the spatially - temporal structure of laser field from a q - switched oscillator depends non - trivially on the oscillator parameters , which effect on the field dynamics at both fundamental and stokes wavelengths . 0.3 , @xmath570.7 , @xmath580 , @xmath59300 cm .
_ b _ : output intensity integrated over @xmath60 ( integral power ; upper plot ) or @xmath4 ( integral fluency ; lower plot).,title="fig:",width=491 ] + fig .
[ fig1 ] , _ a _ shows the contour plot of the output intensity at the fundamental wavelength in absence of an effective srs in an oscillator ( @xmath580 ) .
the intensities are averaged over the cavity period in figs .
[ fig1 ] , _ a _ ; [ fig2 ] , [ fig4 ] , [ fig5 ] , [ fig7 ] , and [ fig8 ] .
both low initial transmission of a saturable absorber and comparatively small radius of curvature of a spherical mirror initiate the multiple pulse dynamics , which is clearly visible in fig .
[ fig1 ] , _ b _ ( upper plot ) showing the output intensity integrated over the transverse spatial coordinate .
the greatest pulse is comparatively long ( @xmath615 ns ) and the narrow `` mode '' ( the radius equals to @xmath61130 @xmath3 m ) has an elongate wing ( lower plot ) .
as will be shown , a non - gaussian transverse shape is typical for the regime under consideration ( see @xcite ) . 0.3 , @xmath570.7 , @xmath580.7 , @xmath59300 cm.,title="fig:",width=491 ] + when a field at the stokes wavelength is locked inside an oscillator owing to non - zero @xmath49 , the dynamics changes .
for the case presented in fig .
[ fig1 ] , the multipulsing disappears ( figs . [ fig2 ] , _ a _ and [ fig3 ] ) and the pulse at the fundamental wavelength shortens substantially ( @xmath611.2 ns , see fig .
[ fig3 ] , upper plot ) @xcite .
the mode at the fundamental wavelength widens and becomes gaussian ( fig .
[ fig3 ] , lower plot ; see also @xcite ) .
simultaneously , the short pulse at the stokes wavelength develops ( figs .
[ fig2 ] , _ b _ and [ fig3 ] ) .
such a pulse is located on the tale of the fundamental pulse ( fig .
[ fig3 ] , gray curve in upper plot ) and has a broad trapezoidal transverse profile ( the mode radius reaches @xmath611.4 mm in fig .
[ fig3 ] , gray curve in lower plot ) .
.,title="fig:",width=302 ] + thus , the transverse distribution excesses substantially the cw mode size , which equals to @xmath61265 @xmath3 m for the spherical mirror under consideration .
such a spatial extra - broadening can be attributed to the saturable absorber saturation , which results in an excitation of high - order spatial frequencies .
a spatial extra - broadening is possible at the fundamental wavelength , as well .
[ fig4 ] , _ a _ demonstrates such an extra - broadening ( up to @xmath610.9 cm ) .
the field covers almost a whole active crystal radius .
it should be noted , that the output cw mode radius equals to @xmath61440 @xmath3 m for the spherical mirror under consideration .
the q - switching `` mode '' has a shape of a dilative ring , which rises and then disappears ( after @xmath6112 ns ) with a gain depletion . in the presence of srs , such a `` mode '' is suppressed owing to an inhomogeneous `` collapsing '' spatially - temporal behavior at both fundamental and stokes wavelengths ( fig .
[ fig4 ] , _ b _ and @xcite ) . 0.3 , @xmath570.5 ; @xmath580 ( _ a _ ) , 0.7 ( _ b _ ) ; @xmath592000 cm.,title="fig:",width=491 ] + an effect of the srs on the q - switching dynamics allows controlling the temporal and spatial profiles of a pulse at both fundamental and stokes wavelengths .
an imbalance between the loss and gain saturation at the fundamental wavelength causing the multiple pulsing can be compensated by the nonlinear loss due to srs . as a result ,
the well - shaped and substantially shortened @xcite pulses with the broad spatial profiles at both wavelengths appear ( figs .
[ fig5],[fig6 ] ) . 0.5 , @xmath570.9 ; @xmath580 ( _ a _ ) , 0.9 ( _ b _ , _ c _ ) ; @xmath591000 cm.,title="fig:",width=529 ] + 0 ( black solid curves ) and 0.9 ( black dotted and gray curves ) .
another parameters correspond to fig .
dotted and gray curves are vertically rescaled for convenience.,title="fig:",width=302 ] + when the oscillator parameters are not optimized , the spatially - temporal structure of the stokes field can be complicated .
for instance , the radial position of the srs spike can depend on the time ( fig .
[ fig7 ] , _ a _ , where the srs spikes move off the optical axis with the pulse evolution ; this effect has been observed experimentally in @xcite ) .
a similar situation is possible also at the fundamental wavelength ( fig .
[ fig4 ] , _ a _ ) . even through
the srs is well - synchronized ( i.e. the srs appears synchronously at the different radial positions ) , the mode profile is inhomogeneous , as a rule ( fig .
[ fig7 ] , _ b _ ) .
the oscillator parameters optimization allows regularizing the spatially - temporal dynamics at both fundamental and stokes wavelengths .
in particular , the spatial profile of the stokes field becomes homogeneous though non - gaussian ( fig .
[ fig8 ] ) .
the numerical analysis reveals , that there exist some optimal sets of parameters ( see table [ t1 ] ) , in particular the optimal curvature of a spherical mirror ( @xmath611000 cm in the case under consideration ) , which provide the regular pulses within a broadest range of parameters ( viz .
@xmath29 , @xmath49 , and @xmath28 ) .
the initial transmission @xmath620.5 results in a sufficiently strong fundamental field , that causes an efficient srs .
the reflectivity at the stokes wavelengths has to be sufficiently high ( @xmath630.7 ) to provide an efficient conversion from the fundamental field to the stokes one .
= 1000 cm , @xmath29=0.9 , @xmath49=0.9 , and @xmath28=0.5.,width=264 ] the resulting pulses at both wavelength are shortened in comparison with a single - wavelength regime and have the smooth and broad transverse spatial profiles .
the last property allows using a larger volume of an active crystal .
the `` mode '' shape is close to the gaussian one for the fundamental field and to the trapezoidal one for the stokes field .
.oscillator parameters providing the regular spatial - temporal pulse profiles at both fundamental and stokes wavelengths .
[ cols="^,^,^,^",options="header " , ]
the model of a q - switched nd : kgw oscillator with a v : yag saturable absorber , which takes into account a transverse spatial field distribution at both fundamental and stokes wavelengths , has been developed . the numerical analysis has demonstrated that an imbalance between the loss and gain saturation causes not only a multiple pulsing but an aberration of an oscillator `` mode '' . when the initial transmission of a saturable absorber is sufficiently low , the spatial extra - broadening of a field develops so that a `` mode '' covers almost whole volume of an active crystal .
manipulations with the beam size on a saturable absorber ( by means of the change of a spherical mirror curvature ) , the initial transmission of an absorber , and the output mirror reflectivity at both fundamental and stokes wavelengths allow controlling the spatially - temporal profiles of the output pulses . to obtain a single pulsing at two - wavelengths with a broad and homogeneous spatial distribution
, the initial transmission of the saturable absorber has to be about of 0.5 , the reflectivity of the output mirror at the fundamental wavelength has to approach 1 ( it can range within 0.5 1 for the optimal curvature of the spherical mirror , which is of @xmath611 m in the case under consideration ) , and the reflectivity at the stokes wavelength has to be about of 0.9 .
the srs causes a temporal squeezing of the pulses at both wavelengths , whereas the spatial profiles are greatly stretched in comparison with the cw - mode size .
when the oscillator parameters are optimized , the transverse spatial profiles are near - gaussian at the fundamental wavelength and trapezoidal at the stokes wavelength .
the obtained results are of interest for development of the passively q - switched raman - active oscillators producing the pulses with durations of about of few nanoseconds and possessing the smooth transverse spatial profiles covering a considerable part of an active crystal .
g.j.spuhler , r.paschotta , r.fluck , b.braun , m.moser , g.zhang , e.gini , u.keller , j. opt .
b 16 ( 1999 ) 376 .
a.m.malyarevich , v.g.savitski , p.v.prokoshin , n.n.posnov , k.v.yumashev , e.raaben , a.a.zhilin , j. opt .
b 19 ( 2002 ) 28 .
i.j.miller , a.j.alcock , j.f.bernard , in : osa proc . advanced solid - state lasers , vol .
13 , 1992 , p.322 .
j.findeisen , h.j.eichler , p.peuser , optics commun . 181 ( 2000 )
a.a.lagatsky , a.abdolvand , n.v.kuleshov , opt .
lett . 25 ( 2000 ) 616 .
a.s.grabtchikov , a.n.kuzmin , v.a.lisinetskii , v.a.orlovich , a.a.demidovich , k.v.yumashev , n.v.kuleshov , h.j.eichler , m.b.danailov , optical meterials 16 ( 2001 ) 349 .
a.s.grabtchikov , a.n.kuzmin , v.a.lisinetskii , v.a.orlovich , and g.i.ryabtsev , appl .
lett . 75 ( 1999 ) 3742 .
a.s.grabtchikov , a.n.kuzmin , v.a.lisinetskii , v.a.orlovich , a.a.demidovich , m.b.danailov , h.j.eichler , a.bednarkiewicz , w.strek , and a.n .
titov , appl .
b 75 ( 2002 ) 795 .
w.chen , y.inagawa , t.omatsu , m.tateda , n.takeuchi , y.usuki , optics commun . 194 ( 2001 ) 401 .
v.l.kalashnikov , optics commun . 218
( 2003 ) 147 .
y.v.loiko , a.a.demidovich , v.v.burakevich , a.p.voitovich , j. opt .
b 22 ( 2005 ) 2450 .
sh.ding , x.zhang , q.wang , j.zhang , sh.wang , y.liu , x.zhang , j. physics d 40 ( 2007 ) 2736 .
a.e.siegman , _ lasers _ ( university science books , 1986 ) .
m.guizar-sicairos , j.c.gutierrez-vega , j. opt .
a * 21 * ( 2004 ) 53 .
v.l.kalashnikov , http://info.tuwien.ac.at/kalashnikov/presentation7/index.htm a.p.shkadarevich , v.i.dashkevich , v.a.orlovich , in : n.s.kazak , p.a.apanasevich , v.v.kabanov , s.n.kurilkina , v.ju.plavskij , s.g.rusov ( eds . ) , proc .
vii international conference on laser physics and optical technologies , v.3 , minsk , belarus , 2008 , p. 477
( in russian ) . | the spatially - temporal model of an all - solid - state passively q - switched oscillator with an active medium providing the stimulated raman scattering is presented .
the model does not presume a gaussian shape of the cylindrically symmetric modes at both fundamental and stokes wavelengths .
it is found , that the highly - nontrivial spatially - temporal dynamics can be regularized by the optimal choice of the oscillator parameters , viz .
initial transmission of a saturable absorber , curvature of a spherical mirror , and output mirror transmission at the fundamental and stokes wavelengths . as a result
, the pulse can be substantially temporally squeezed and spatially broadened at both fundamental and stokes wavelengths .
q - switching ; solid - state lasers ; stimulated raman scattering ; spatially - temporal dynamics |
flat spacetimes are not as sophisticated or realistic as their curved counterparts , but they are still worth studying as they can provide insight to general relativity . for example , there are the milne and rindler universes ( see @xcite ) , and , in a quite different flavor , there is the
twin paradox on compact space " ( see e.g. @xcite-@xcite ) .
the paper by bansal , laing , and sriharan @xcite concerns a flat spacetime @xmath0 with topology @xmath1 .
this spacetime can be constructed from standard minkowski spacetime @xmath2 by declaring points with coordinates @xmath3 _ equivalent _ to points with coordinates @xmath4 , where @xmath5 measures the length of the compact circular " spatial dimension and @xmath6 .
the resulting manifold @xmath0 inherits the flat minkowskian metric @xmath7 from @xmath2 .
each event ( or point ) in @xmath0 corresponds to an equivalence class of points @xmath8_{n\in\mathbb{z}}$ ] in @xmath2 .
hence , each point in @xmath0 can be uniquely expressed by coordinates @xmath3 such that @xmath9 . in other words
, one has a quotient map which projects each point @xmath3 of @xmath2 to the point of @xmath0 given by coordinates @xmath10 .
the worldline of an inertial observer in @xmath0 can be regarded as the projection of an inertial observer in @xmath2 the equivalence relation @xmath11 , used above to describe the topology of @xmath0 , is coordinate dependent .
the lorentz transformation given by the matrix @xmath12= \left [ \begin{array}{rrrr } ( \gamma-1)(u^1)^2 + 1 & ( \gamma-1)u^1u^2 & ( \gamma-1)u^1u^3 & -\beta\gamma u^1 \\ ( \gamma-1)u^1u^2 & ( \gamma-1)(u^2)^2 + 1 & ( \gamma-1)u^2u^3&-\beta\gamma u^2\\ ( \gamma-1)u^1u^3&(\gamma-1)u^2u^3&(\gamma-1)(u^3)^2 + 1&-\beta\gamma u^3\\ -\beta\gamma u^1&-\beta\gamma u^2&-\beta\gamma u^3&\gamma \end{array}\right],\ ] ] where @xmath13 , describes the transformation from the unprimed " frame @xmath3 to a primed " frame @xmath14 in @xmath2 which moves at a uniform velocity @xmath15 , with @xmath16 such that @xmath17 , relative to the unprimed frame ( see e.g. @xcite , page 69 ) . for such an observer , the equivalence relation for describing the topology of @xmath0 becomes , for any @xmath6 , @xmath18 whence , a general boosted observer identifies points not only across space , but also across time ( if @xmath19 ) . if this time shift "
@xmath20 is nonzero , then there exist two separated spatial positions , which would normally be regarded as simultaneous by the einstein convention , that are assigned two different times by this observer .
hence , such an observer ( or more precisely , the projection of such an observer down to @xmath0 ) can not set up a global " system of einstein - synchronized clocks encompassing the entire universe . on the other hand ,
if @xmath21 , then the observer in minkowski spacetime identifies ( what she regards as ) simultaneous spatial positions through space only , preserving the assignment of a unique time to each event .
hence , such an observer has the special privilege of being able to set up a global system of einstein - synchronized clocks in @xmath0 .
we will call these specially privileged observers , * privileged*. the upshot is that @xmath0 does not have a unique privileged frame but instead has infinitely many privileged frames ( those described by setting either @xmath22 or @xmath23 , so that @xmath21 ) .
these privileged frames exhibit inertial motions with respect to each other .
this is unlike the flat spacetime with topology @xmath24 studied in @xcite .
flat @xmath24 spacetime has a _ unique _
privileged frame , which by virtue of its uniqueness can be regarded as an absolute rest frame . " here is an elementary physical argument for the non - uniqueness of privileged frames in @xmath0 .
an inertial observer can determine whether or not she is privileged by sending out a flash of light from her position .
this flash of light will spread out in all directions , including the compact dimension , at the speed of light . by looking along the compact dimension
, the observer will eventually be able to see the light flash that she had previously initiated .
if she is moving along a non - compact dimension , but _ not _ moving along the compact dimension , then she will see the light flash in both directions along the compact dimension at the same time , and hence she will discover that she is privileged .
if she _ is _ moving along the compact dimension , then she will see her previous light flash in one direction before seeing it in the opposite direction , and hence she will learn that she is not privileged .
( see figure [ cylinders3 ] ) . and
@xmath25 receive the flash from two directions at the same time ( according to their own respective observations ) .
observer @xmath26 has already received a flash from one direction and will soon receive a flash from the other .
the conclusion is that observers @xmath27 and @xmath25 are privileged , but observer @xmath26 is not.,width=384 ] this raises an issue . by using simple physical experiments involving flashes of light
, inertial observers can learn whether or not they are privileged .
hence the principle of relativity , strictly applied , breaks down in this model .
however , the principle of relativity does hold _ locally _ in @xmath0 , as we show in section 3 after the locality concept is sufficiently clarified .
thus , inertial observers in @xmath0 can not use _ local _ experiments to determine whether or not they are privileged .
the authors of @xcite propose , to the contrary , that inertial observers in @xmath0 _ can _ use a local experiment ( involving electrostatics ) to determine whether or not they are privileged . in section 3
we will explain why their experiment is not really local , but first we need to describe the electrostatic field in @xmath0 .
our analysis corrects errors made by @xcite on the matter .
in @xcite , the analysis is implicitly constrained to the case where the primed " frame exhibits inertial motion with respect to the unprimed " frame at a velocity @xmath15 , where @xmath28 .
in such a frame , the equivalence relation describing the topology of @xmath0 becomes , for any @xmath6 , @xmath29 note that in this frame , the length of the compact spatial dimension is effectively @xmath30 .
this is somewhat larger than the value found by a privileged observer . with respect to the unprimed frame ,
the effective size of the compact dimension is @xmath31 .
the physical meaning of @xmath32 is that an inertial observer in @xmath0 can not extend her coordinate system farther than @xmath32 in the compact direction , if her coordinate system synchronizes clocks by the einstein procedure .
note that @xmath33 , with equality holding if and only if one is in a privileged frame . ]
assume that at the spatial origin of such a frame , there sits a stationary point - charge .
it has existed there for all past eternity and will continue to exist there for all future eternity .
its electric charge is , always has been , and always will be , an absolute nonzero constant @xmath34 .
there are no other electric charges or magnetic fields in @xmath0 .
since @xmath0 has the topology @xmath35 , one expects that the electrostatic field of such a point - charge should deviate from coulomb s inverse - square law .
indeed , the @xmath36-component of the electrostatic field generated by the charge should vanish at the antipodal " point on the @xmath36-axis with coordinate @xmath37 .
so the electrostatic field in @xmath0 should decrease _ faster _ than the inverse - square of the distance along the @xmath36-axis . as explained in @xcite , the electrostatic field at the spatial point @xmath38 in @xmath0 in such a reference frame
is given by @xmath39^\frac{3}{2}},\ ] ] where we have dropped the primes " and written @xmath40 in place of @xmath32 for simplicity .
consider the @xmath41-component of the electric field along the @xmath41-axis ( where @xmath42 and @xmath43 vanish ) .
it is given by @xmath44^\frac{3}{2}}.\ ] ] this function is periodic with period @xmath40 , as one would expect since the @xmath41-dimension is periodic with length @xmath40 ( for an analytical proof , see proposition 1 in the appendix ) . note that , if @xmath45 , @xmath46^\frac{3}{2}}&=&\frac{1}{x^2}+\sum_{n=1}^\infty\left(\frac{1}{(x+nl)^2}-\frac{1}{(x - nl)^2}\right ) \nonumber \\ & = & \frac{1}{x^2}+\frac{1}{l^2}\sum_{n=1}^\infty\left(\frac{1}{(x / l+n)^2}-\frac{1}{(x / l - n)^2}\right ) . \nonumber\end{aligned}\ ] ] defining a function @xmath47 through the series@xmath48which converges absolutely for @xmath49 ( proposition 2 in the appendix ) , we can write , for @xmath45 , @xmath50 note that @xmath51 . hence , @xmath52 , as expected .
moreover , one verifies that the series defining @xmath53 converges uniformly over compact subsets of the unit disc @xmath54 ( lemma 1 in the appendix ) .
it follows ( see @xcite , page 214 ) , that @xmath47 is holomorphic on @xmath55 and all the derivatives of @xmath47 can be calculated by differentiating behind the summation symbol .
one shows by mathematical induction that @xmath56.\ ] ] hence @xmath57 vanishes if @xmath58 is even and @xmath59 if @xmath58 is odd , where @xmath60 is the riemann zeta function ( see e.g. , @xcite , page 212 ) .
we therefore have a power series expansion @xmath61 it can be verified that the radius of convergence for this power series is 1 ( proposition 3 in the appendix ) . in other words , the power series is valid on @xmath55 .
consequently , for @xmath45 , @xmath62 so we get that the first - order correction term in @xmath0 to the usual coloumb law is @xmath63 .
the absolute fractional difference from the usual electrostatic field ( along the @xmath41-axis ) is therefore approximated by @xmath64 the fact that @xmath47 is an odd function with @xmath65 , and not , as reported in @xcite , an even function which is constant to first - order , is anticipated by the following physical argument .
let @xmath47 denote the correction function to the electrostatic field along the @xmath41-axis .
note that we are now using the name
@xmath47 " somewhat abusively , since previously @xmath66 is the correction to the electrostatic field along the @xmath41-axis where @xmath67 .
for the present argument , we are taking @xmath68 and @xmath69where @xmath70 .
physically , @xmath71 can be interpreted as the strength of an electrostatic field at the point @xmath72 on the @xmath41-axis ( in @xmath73 ) generated by an infinite sequence of point - charges , each of charge @xmath34 , individually glued down ( so that they do nt move ) onto the @xmath41-axis at the coordinates @xmath74 . by symmetry ,
the electrostatic fields of the individual charges cancel out near @xmath75 .
in other words , @xmath76 . moreover , since the electrostatic field on the @xmath41-axis at the points @xmath77 and @xmath78 ( with @xmath70 ) should have the same magnitude but opposite direction , one expects that @xmath79 . that is , @xmath47 is an odd function .
in @xcite the claim is made that in the spacetime @xmath0 a local " experiment , involving the electrostatic field of a point - charge , can be used to determine whether or not an inertial observer is privileged .
the idea is to note that the law for electrostatics in @xmath0 depends on the ( effective ) size of the compact dimension , as we have found in section 2 .
as noted in section 2 , the compact dimension has length @xmath40 in a privileged frame but has a somewhat larger length in a non - privileged frame .
the authors of @xcite conclude from these facts that an inertial observer in @xmath0 can determine ( or at least estimate ) the effective size of the compact dimension in their frame by making
nearby " measurements of the electrostatic field strength generated by a stationary point - charge in their frame .
the results found in section 2 support the idea that this can be done , at least in principle . but rather the more immediate question of whether the experiment proposed by @xcite is truly local , even if it is done in only one frame .
] however , we do not agree with the claim that this is a truly local " experiment .
the principle of relativity , at least in general relativity , says that the _ local _ laws of physics do not distinguish between two inertial observers .
hence , there is no _ local _ experiment that an observer can perform in order to determine whether or not she is privileged .
therefore , the local " experiment proposed by @xcite can not be called
local " if we wish to retain the principle of relativity .
yet , a dogmatic commitment to the principle of relativity does nothing to explain why the experiment is not a local one . at a more basic level
, one can argue as follows .
we start by investigating the meaning of the word local . " in the present context one can distinguish between two notions of locality : * geometrical locality * and * topological locality * ( compare @xcite ) .
we shall say that a physical system is local in the geometrical sense if it exists wholly within an infinitesimal region " of the spacetime manifold . on the other hand
, we shall say that a physical system is local in the topological sense if it exists wholly within a region we call a patch " of the manifold . since an infinitesimal region ( whatever that is ) can be contained within a single patch of @xmath0 , it follows that the topological notion of locality generalizes the geometrical notion .
thus , if a system is not local in the topological sense , it can not be local in the geometrical sense either .
it is sufficient for our purposes to note that a properly local experiment necessarily involves only a physical system which can be completely described within one patch of the spacetime manifold .
we can apply this notion to see why the experiment described in the introduction of this note , the experiment involving a light flash , is not local .
this thought experiment concerns a physical system involving light signals which traverse the entire compact dimension of @xmath0 .
hence , at least _ two _ patches of @xmath0 would be needed in order to completely describe the experiment . in a certain nave sense
, this experiment only requires the observer to make a few local time measurements on his or her worldline ; whence we notice that even if an experiment wholly relies on measurements which navely appear to be local , " it does not follow that the whole experiment itself really _ is _ local .
the same argument carries over to the local " thought experiment proposed by @xcite .
their experiment only requires the observer to make apparently local " ( or nearby " ) measurements of an electrostatic field , but this does not mean that the experiment is truly local . indeed , a more _ complete _ description of the experiment would account for the underlying physical systems which influence each measurement .
for example , consider the question : what produces the local " electrostatic field and what determines its structure ?
the producer of the field is the point - charge source ( in the sense that , in this thought experiment , there is no field without the point - charge ) , and the structure of the field ( at any point where it is defined ) depends on global " topological features of the manifold ( a fact which is not overlooked in @xcite ) .
hence , the experiment proposed by @xcite , if described more completely , actually involves a physical system which includes the point - charge source together with an electrostatic field generated by the charge , which wraps all the way around the compact dimension .
this physical system can not be described within a single patch of the manifold .
therefore , the experiment is not local .
indeed , the principle of relativity _ must _ hold locally in @xmath0 .
a local experiment is a physical system which can be completely contained in a single patch of @xmath0 .
a patch of @xmath0 is essentially indistinguishable from a region of minkowski spacetime @xmath2 , in which the principle of relativity certainly holds .
this argument is independent of the topology of @xmath0 .
( see also @xcite . )
the purpose of this appendix is to fill in the mathematical gaps left in section 2 . throughout
, @xmath40 is a positive constant . let @xmath82 .
according to the interpretation of a doubly infinite sum ( e.g. , @xcite , page 184 ) , we have @xmath46^\frac{3}{2}}&:= & \sum_{n=0}^\infty\frac{x+nl}{[(x+nl)^2]^\frac{3}{2}}+\sum_{n=1}^\infty\frac{x - nl}{[(x - nl)^2]^\frac{3}{2 } } \nonumber \\ & = & \sum_{n=0}^\infty\frac{x+nl}{|x+nl|^3}+\sum_{n=1}^\infty\frac{x - nl}{|x - nl|^3}. \nonumber\end{aligned}\ ] ] note that @xmath83 and @xmath84 . for @xmath85 , one verifies that both @xmath86 and @xmath87 converge by the integral test " ( see @xcite , page 139 ) .
hence , it follows that @xmath88^\frac{3}{2}}$ ] converges absolutely .
moreover , we can write @xmath89^\frac{3}{2}}=\frac{x}{|x|^3}+\sum_{n=1}^\infty\left(\frac{x+nl}{|x+nl|^3}+\frac{x - nl}{|x - nl|^3}\right).\ ] ] for @xmath94 and @xmath95 , we have that @xmath96
choosing @xmath97 , one finds that the series @xmath98 converges by the integral test .
whence , the series @xmath99 converges absolutely for @xmath94 , by the so - called comparison test " ( @xcite , page 60 ) .
it suffices to show that the series converges uniformly on @xmath101 , where @xmath102 .
we have that , for @xmath103 , @xmath104 note that , for @xmath105 , we have that @xmath106 .
hence , @xmath107 , and we find that @xmath108 the series @xmath109 converges by the integral test .
it now follows , by the weierstrass @xmath0-test , " ( @xcite , page 37 ) that @xmath93 converges uniformly over @xmath110 .
since we have verified that @xmath113 $ ] converges uniformly over compact subsets of @xmath55 ( lemma 1 ) , we have that the sequence @xmath114 , where @xmath115,\ ] ] converges uniformly over compact subsets of @xmath55 to @xmath47 .
note also that each @xmath116 is holomorphic on @xmath55 . since @xmath119\nonumber \\ & = & \lim_{n\rightarrow\infty}\sum_{n=1}^n\frac{d^k}{dz^k}\left[(z+n)^{-2}-(z - n)^{-2}\right ] \nonumber \\ & = & \sum_{n=1}^\infty\frac{d^k}{dz^k}\left[(z+n)^{-2}-(z - n)^{-2}\right],\nonumber\end{aligned}\ ] ] we have that the derivatives of @xmath47 can all be calculated by differentiating behind the summation symbol . to complete the proof of this lemma
, we show that @xmath120.\ ] ] this formula can be established by mathematical induction .
we have that @xmath121 , and according to the formula , @xmath122\nonumber \\ & = & \sum_{n=1}^\infty\left[(z+n)^{-2}-(z - n)^{-2}\right]\nonumber \\ & = & \varphi(z).\nonumber\end{aligned}\ ] ] for induction , suppose @xmath123.$ ] then , since we can differentiate behind the summation symbol , @xmath124\nonumber \\ & = & ( -1)^{m+1}(m+2)!\sum_{n=1}^\infty\left[(z+n)^{-2-(m+1)}-(z - n)^{-2-(m+1)}\right ] .
\nonumber\end{aligned}\ ] ] thus , by induction , the lemma holds .
rudin ( page 214 in @xcite ) remarks that for functions on the real line , it is possible for sequences of infinitely differentiable functions to converge uniformly to nowhere differentiable functions .
this is the reason why we find it desirable to consider @xmath47 as a complex function .
it keeps us on the safe side . in the proof of lemma 2
, it was noted that the function @xmath47 is holomorphic on @xmath55 .
hence , @xmath47 is representable by power series in @xmath55 ( see , e.g. , @xcite page 207 , theorem 10.16 ) .
this means that for every point @xmath128 , there is an open disc @xmath129 and a corresponding power series of the form @xmath130 which converges to @xmath53 for all @xmath131 ( @xcite , page 198 ) . by the corollary to @xcite s theorem 10.16 ( appearing in @xcite on page 199 ) , the constants @xmath132 are unique and are in fact given by @xmath133 the radius @xmath134 can not exceed a certain number @xmath135 $ ] called the radius of convergence " for the series .
the radius of convergence @xmath136 is given by
hadamard s formula " ( @xcite page 38 ) @xmath137 the power series @xmath138 converges absolutely and uniformly in @xmath139 for every @xmath140 and diverges if @xmath141 ( see @xcite page 198 )
. we are concerned with the power series representation for @xmath47 centered at @xmath142 . by lemma 2
, it follows that the power series representation for @xmath47 centered at @xmath143 is @xmath144 the remaining issue is to find the radius of convergence @xmath136 for this series . by lemmas 2 , 3 , and hadamard s formula
, it follows that @xmath145 our claim then , is that @xmath146 one has that @xmath147 for any positive constant @xmath148 .
so @xmath149 . to see that @xmath150 , set @xmath151 ( at this point we are adapting an argument from page 39 in @xcite )
. then we can write @xmath152 .
note that @xmath153 is positive .
the binomial theorem gives @xmath154 therefore , provided @xmath155 , we get that @xmath156 .
hence @xmath157 as @xmath158 and this proves that @xmath150 it remains to be shown that @xmath159 .
note that it suffices to show that @xmath160 ( @xmath161 ) since it would then follow that @xmath162 for sufficiently large @xmath58 , and hence @xmath159 by the squeeze theorem " of freshman calculus .
notice that @xmath166 as @xmath158 ( since @xmath167 ) and @xmath168 for all @xmath58 .
thus , by the lebesgue dominated convergence theorem " ( @xcite page 26 ) , it follows that @xmath169 where @xmath170 and @xmath171 is the counting measure .
in other words , @xmath172 as claimed . | we study a flat spacetime with a compact spatial dimension and dispute the claim of bansal , laing , and sriharan that a local experiment can determine whether or not an inertial observer is privileged " in this spacetime . |
general relativity coupled to maxwell fields admits static charged black hole solutions in spacetime dimensions @xmath7 and higher . the spacetimes are warped products @xmath8 of a two dimensional lorentzian
orbit manifold " @xmath9 with line element @xmath10 and an @xmath11 dimensional riemannian `` horizon manifold '' @xmath12 with metric @xmath13 ( see , e.g. , @xcite ) : @xmath14 in four dimensions , the solution with @xmath15 horizon is the reissner - nordstrm black hole .
if we use the standard angular coordinates @xmath16 and static coordinates @xmath17 for the orbit manifold , the reissner - nordstrm metric is given by @xmath18 where the norm @xmath19 of the killing vector @xmath20 in ( [ rn ] ) is @xmath21 in ( [ f ] ) , @xmath22 is the cosmological constant , and @xmath23 and @xmath24 are constants of integration that correspond to mass and charge respectively .
the metric ( [ rn ] ) together with the maxwell field @xmath25 solves the einstein - maxwell field equations @xmath26 } = 0,\\ \label{max2 } & \nabla^{\b } f_{\a \b } = 0,\end{aligned}\ ] ] where @xmath27 note that , since @xmath28 is traceless , ( [ efe ] ) is equivalent to @xmath29 we assume @xmath30 and focus in the black hole cases , which are those for which there is an outer static ( @xmath31 ) region , that is , either @xmath32 and @xmath33 , or @xmath34 and @xmath35 . here
the event and cosmological horizons @xmath36 and @xmath37 are simple zeros of the quartic polynomial @xmath38 if the black hole is non extremal .
the range of values of @xmath39 and @xmath22 giving black holes can be found in appendix a of @xcite .
+ we are interested in proving the non - modal linear stability of the outer static region of the solution ( [ rn])-([max ] ) of the field equations ( [ efe])-([efe2 ] ) . by this we mean @xcite @xcite , showing that : a. there are gauge invariant ( both in the maxwell and linear gravity sense ) scalar fields @xmath40 that contain the same information as the perturbation @xmath41 of the electromagnetic field and the gauge class @xmath42 $ ] of the metric perturbation @xmath43 , and measure the distortion of the geometry and the maxwell field . by `` contain the same information '' we mean that @xmath44 in a given gauge and @xmath45 can be obtained by applying some injective linear functional on the fields @xmath46 .
b. the fields @xmath46 are pointwise bounded on the outer static region by constants that depends on the initial data of the perturbation on a cauchy surface .
the perturbed metric and maxwell fields can be expanded in series involving rank 0,1 , and 2 eigentensor fields of the horizon manifold laplace - beltrami ( lb ) operator , with `` coefficients '' that are tensor fields in the orbit space @xmath47 ; this is the mode expansion of @xmath44 and @xmath45 @xcite .
the linearized einstein - maxwell equations ( leme ) do not mix modes . a master scalar field @xmath48 can be extracted for each mode such that the leme reduce to an infinite set of _ scalar wave equations on _ @xmath47 , one for each master scalar field .
this was proved in four dimensional general relativity in the seminal black hole stability papers @xcite @xcite @xcite and in higher dimensions more recently by kodama and ishibashi ( see , eg.g .
@xcite and @xcite ) .
prior notions of linear stability are based on the boundedness of the master fields on the orbit manifold @xmath47 , we call this _ modal ( linear ) stability_. in the case of four dimensional asymptotically flat charged black holes the modal linear stability was proved by zerilli and moncrief in the series of articles @xcite ( see also @xcite ) + the limitations of the modal linear stability notion are explained in @xcite and @xcite , where a non modal stability concept based on i ) and ii ) above was proved to hold for the schwarzschild and schwarzschild de sitter black holes . in these papers the fields @xmath46 in i ) are gauge invariant combinations of perturbations of scalars made out of contractions of the weyl tensor , its dual , and its first covariant derivative .
+ for einstein - maxwell black holes the extra degrees of freedom coming from the maxwell field have to be accounted for .
perturbations naturally split into two decoupled types : odd and even ( section [ rwz ] ) . in this paper
we prove the non modal linear stability of the reissner - nordstrm black hole under odd perturbations .
the fields @xmath46 that fulfill the requirements i ) and ii ) above are the first order perturbation of the scalars obtained by contracting the maxwell and weyl tensors with their hodge duals : @xmath0 and @xmath49 .
these fields are shown to satisfy a coupled system of wave equations in the reissner - nordstrm background , and this fact is used to prove their pointwise boundedness on the outer static region .
we defer to future work the treatment of even perturbations .
assume that @xmath50 is a smooth one - parameter set of solutions of the einstein - maxwell equations ( [ efe2])-([efe1 ] ) , with @xmath51 the reissner - nordstrm fields ( [ rn])-([max ] ) , then the _ perturbation fields _
@xmath52 satisfy the linearized einstein - maxwell equations ( leme ) : @xmath53,\\ \label{2 } & \frac{d}{d \ve}\bigg|_{\ve=0 } \p_{[\a } f({\ve})_{\b \g ] } = 0,\\ \label{3 } & \frac{d}{d \ve}\bigg|_{\ve=0 } \nabla_{\b } f({\ve})^{\a \b } = 0.\end{aligned}\ ] ] as in equation ( [ warped ] ) , we adopt the notation in @xcite and use lower case indexes @xmath54 for tensors on the orbit manifold @xmath9 , upper case indexes @xmath55 for tensors on @xmath15 , and greek indexes for space - time tensors , and follow the additional convention in @xcite that @xmath56 tensor fields _ introduced _ with a lower @xmath15 index ( say @xmath57 ) and _ then shown with an upper @xmath15 index _ are assumed to have been acted on _ with the unit @xmath15 metric inverse @xmath58 _ , ( i.e. , in our example , @xmath59 ) , and similarly with upper @xmath15 indexes moving down .
this has to be kept in mind to avoid wrong @xmath60 factors in the equations . @xmath61 and @xmath62 are the covariant derivative , volume form ( any chosen orientation ) and metric inverse for the @xmath47 orbit space ; @xmath63 and @xmath64 are the covariant derivative and volume form @xmath65 on the unit sphere . as an example , the laplacian on scalar fields can be written in terms of the differential operators @xmath66 and @xmath63 as @xmath67 the linearized field equations , ( [ lf ] ) and ( [ 2 ] ) imply that locally there exists a vector potential @xmath68 such that @xmath69 the linear fields entering the leme are @xmath44 and @xmath68 . under the index convention ( [ index ] ) the covector field @xmath68
is written @xmath70 from the @xmath15 viewpoint @xmath68 contains two scalar fields @xmath71 and a covector field @xmath72 .
using proposition 2.1 in @xcite and the fact that the first betti number of @xmath15 is zero ( which implies that divergence free @xmath15 covectors are of the form @xmath73 , with @xmath74 an @xmath15 scalar field ) , we can write @xmath75 , thus @xmath76 the scalar fields @xmath77 are unique if they are required to belong to @xmath78 @xcite , where @xmath79 is the space of square integrable functions on @xmath15 orthogonal to the @xmath80 eigenspaces of the lb operator , and @xmath81 labels the lb scalar field eigenvalue @xmath82 . the plus ( even ) and minus ( odd ) signs on tensor fields refer to the way they transform when pull backed by the antipodal map @xmath74 on @xmath15 @xcite .
+ a symmetric tensor field @xmath83 , @xmath84 such as the perturbations of the metric , the einstein and the energy momentum tensor fields , contains three @xmath15 scalar fields @xmath85 , two @xmath15 covector fields @xmath86 and a symmetric tensor field @xmath87 .
the @xmath15 covectors can be decomposed as in ( [ adecomp ] ) @xmath88 where again @xmath89 are unique if their components are in @xmath78 . from proposition 2.2 in @xcite and
the fact that there are no transverse traceless symmetric rank two tensor fields on @xmath15 , follows that @xmath90 the fields @xmath91 are unique if required to belong to @xmath92 @xcite . in this way
, the symmetric field @xmath93 is replaced by two sets of fields , even ( + ) and odd ( - ) : @xmath94 if we decompose the linearized symmetric tensor fields @xmath44 , @xmath95 and @xmath96 as in ( [ std1])-([std4 ] ) , we get the following sets of even and odd fields : @xmath97 group theoretical arguments ( refer to section 2 of @xcite ) indicate that the leme involving the even fields in ( [ adecomp ] ) ( [ hdecomp])-([tdecomp ] ) decouple from those involving the odd fields , so we can switch off one sector and study purely odd or even perturbations .
odd perturbations are the subject of this paper
. + we will find it useful to introduce the square angular momentum operator @xmath98 where @xmath99 and @xmath100 are @xmath15 ( and thus spacetime ) killing vector fields corresponding to rotations around orthogonal axis in @xmath101 , with maximum orbit orbit lengths set to @xmath102 ( e.g. , @xmath103 ) . on @xmath15 scalar fields
the operator @xmath104 agrees with the @xmath15 lb operator @xmath105 , but these two operators act differently on higher rank tensors .
a key property of @xmath104 is that it commutes with @xmath106 , @xmath66 and @xmath63 , this follows from @xmath107 = 0 = [ \hd_a , \pounds_{j_k } ] = [ \td_a , \pounds_{j_k}]$ ] .
the modal decomposition consists in expanding the @xmath15 scalars in ( [ adecomp ] ) ( [ hdecomp])-([tdecomp ] ) in a real basis of spherical harmonics of @xmath15 , which are eigenfields of @xmath104 with eigenvalues @xmath108 , the eigenspaces being of of dimension @xmath109 .
the differential operators that give a symmetric tensor @xmath93 or a covector @xmath68 in terms of these @xmath15 scalars commute with @xmath104 .
thus , if the @xmath15 scalar fields in ( [ hdecomp ] ) and ( [ adecomp ] ) lie on the @xmath81 eigenspace , then @xmath110 will all be eigentensors of @xmath104 with eigenvalue @xmath108 , i.e. , different modes stay unmixed .
the distinction between even and odd modes can now be stated in a precise way : if @xmath111 is a covector ( [ adecomp ] ) or symmetric field @xmath93 ( [ std1])-([std4 ] ) of a given parity , made out of scalars of harmonic numbers ( @xmath112 , then @xmath113 and @xmath114 .
+ we will assume that @xmath115 to @xmath78 and @xmath91 to @xmath92 , then the linear operators @xmath116 in ( [ adecomp ] ) , and @xmath117 in ( [ std3 ] ) are injective @xcite .
consequently , the odd sector leme ( [ 1 ] ) are equivalent to @xmath118 odd perturbations are those for which the plus fields in ( [ adecomp ] ) and ( [ hdecomp ] ) are zero : @xmath119 @xmath120 gauge transformations of the maxwell field are of the form @xmath121 and therefore affect only the even piece of the vector potential ( [ adecomp ] ) . under a coordinate gauge transformation ( infinitesimal diffeomorphism ) along the odd vector field @xmath122 , @xmath123 , @xmath44 and @xmath124 transform into the physically equivalent fields : @xmath125 we denote @xmath126 the set of odd solutions @xmath127 of the leme ( [ 1])-([3 ] ) mod the equivalence relation @xmath128 above , that is , if @xmath42 $ ] denotes equivalence class under the transformation ( [ prime ] ) , then @xmath129 , \f_{\a \b } ) \ ; | \ ; ( h_{\a \b } , \f_{\a \b } ) \ ; \text { is an odd solution of ( \ref{1})-(\ref{3 } ) } \}.\ ] ] the transformation ( [ prime ] ) is equivalent to @xmath130 where @xmath131 is the projection of @xmath132 onto @xmath92 .
this implies that a. the @xmath47 1-form @xmath133 and scalar field @xmath134 defined by @xmath135 are gauge invariant ; b. there exists a gauge where the metric perturbation looks like @xmath44 in ( [ pert1 ] ) with @xmath136 and @xmath137 replaced with @xmath138 .
this is the well known regge - wheeler ( rw ) gauge . from now on we work in the rw gauge ,
then we set @xmath136 in ( [ pert1 ] ) and replace @xmath137 with @xmath139 .
note that the absolute value @xmath140 of the determinant of the metric in this gauge agrees ( to linear order ) with the absolute value @xmath141 of the unperturbed metric determinant , then @xmath142 to linear order the inverse metric is @xmath143 this is used to raise the indexes of the perturbed maxwell field @xmath144 .
the result is @xmath145 taking advantage of ( [ det ] ) we find that the linearized maxwell equation ( [ 3 ] ) can be written @xmath146 and using @xmath147 , we find that the @xmath148 equation above is trivial whereas the @xmath149 equation gives @xmath150,\ ] ] which , since @xmath151 , is equivalent to @xmath152 the linearized einstein s equations ( [ 1 ] ) are equivalent to the set ( [ ea ] ) and ( [ e ] ) .
after a lengthly calculation we find @xmath153 we also find that @xmath154 and @xmath155 since @xmath136 , equations ( [ e ] ) and ( [ g - t- ] ) give @xmath157 .
the solution of this equation is @xmath158 for some potential @xmath159 , defined up to the sum of a function of @xmath160 : @xmath161 inserting ( [ ta ] ) , ( [ ga ] ) and ( [ ha ] ) into the projection onto @xmath92 of the linearized einstein equation ( [ ea ] ) gives @xmath162 + \left ( \td^c \td_c r^2 + 2 r^2 \lambda + 2 \frac{q^2}{r^2 } \right ) \e_a{}^b \td_b z -4q \e_a{}^b \td_b a^{>1 } = 0\ ] ] the fact that @xmath163 makes it possible to pull operator @xmath164 to the left in the above equation .
since the kernel of @xmath164 acting on @xmath47-scalar fields are the @xmath47-constants ( i.e. , functions of @xmath165 ) , we can lift @xmath164 from this equation and get @xmath166 we now use the freedom ( [ zf ] ) and choose @xmath167 to cancel @xmath168 .
this is possible since the operator @xmath169 is invertible in @xmath92 .
this choice of @xmath159 is equivalent to setting @xmath170 in ( [ eq1d2 ] ) .
the resulting equation is equivalent the four dimensional wave equation @xmath171 where @xmath172 the equation obtained after inserting ( [ ha ] ) into the projection onto @xmath92 of the linearized maxwell equation ( [ divf ] ) and then using ( [ eq1d2 ] ) , @xmath173 also admits the form of a four dimensional wave equation linking @xmath174 and @xmath46 above : @xmath175 here we used the facts that on scalar fields @xmath176 and @xmath177 .
+ we conclude that the odd sector @xmath156 leme are entirely equivalent to the system of ( four dimensional ) wave equations ( [ master ] ) and ( [ master2 ] ) coupling the fields @xmath46 and @xmath174 , which are potentials from which the metric perturbation in the rw gauge and the electromagnetic field perturbations are given by equations ( [ pert1 ] ) with @xmath178 and @xmath136 , ( [ ha ] ) and ( [ wfi0 ] ) . the projection of the linearized maxwell equation ( [ divf1 ] ) onto the three dimensional @xmath179 subspace of @xmath180 is @xmath181 and the projection of the leme ( [ ea ] ) , using ( [ ta ] ) and ( [ ga ] ) is @xmath182=0.\ ] ] we will show that the solutions of this system of equations correspond to perturbations along the stationary kerr - newman ( anti ) de sitter ( kn(a)ds ) mono - parametric family of solutions of the einstein - maxwel equations , with the angular momentum playing the role of @xmath183 in ( [ lf ] ) .
+ consider the kn(a)ds black hole solution with mass @xmath23 and angular momentum @xmath184 along the @xmath185 axis in boyer lindquist coordinates .
if we taylor expand the metric around @xmath186 we obtain @xmath187 where @xmath188 is the reissner - nordstrm metric ( [ rn])-([f ] ) , @xmath189 the remaining components being trivial .
we recognize that @xmath44 is a perturbation in the regee - wheeler gauge , that is , of the form ( [ pert1 ] ) .
since @xmath190 and @xmath191 , the only solution of ( [ hti])-([hfi ] ) and @xmath192 is @xmath193 this is an @xmath179 perturbation , as @xmath194 is the linear span of the orthonormal basis @xmath195 if we chose an element of @xmath196 other than @xmath197 in ( [ ha-1 ] ) , the resulting perturbation , @xmath198 is that corresponding to an an infinitesimal rotation with angular momentum components @xmath199 .
+ the nonzero components of the maxwell vector potential @xmath68 for the electromagnetic field @xmath200 of the kn(a)ds black hole are @xmath201 which is an odd perturbation with @xmath202 this generalizes to @xmath203 it now is easily verified that @xmath139 in ( [ ha-1-g ] ) together with @xmath134 in ( [ a-1-g ] ) are , mod the gauge transformations ( [ gauge ] ) , the most general solutions of the @xmath179 leme ( [ divf - l=1])-([leme - l=1 ] ) .
+ for future reference , we note that for the kn(a)ds spacetime @xmath204
the non - modal stability concept outlined in items i ) and ii ) in the introduction requires analyzing the effects of the perturbation on gauge independent scalar fields on the spacetime manifold . from the results of the previous section
follows that the set @xmath126 of odd perturbations mod gauge transformations , equation ( [ l- ] ) , can be parametrized by the gauge invariant fields @xmath205 subjet to equations ( [ master])-([master2 ] ) and the constants @xmath206 , that is @xmath207 there is a distinction , however between the @xmath179 constants , which are the components of an infinitesimal angular momentum , corresponding to perturbations within the kerr - newman ( a)ds family and therefore physically meaningful , and the scalar fields @xmath46 and @xmath174 , which are convenient potentials to solve the @xmath156 leme without any obvious physical interpretation .
+ we prove in section [ invs ] that there are two gauge invariant , physically meaningful scalar fields , @xmath208 and @xmath5 , directly associated to the effects of the perturbation on the curvature and on the strength of the maxwell field , that contain the same information as the gauge class @xmath42 $ ] and the perturbed electromagnetic field @xmath45 for generic odd perturbations ( that is , containing @xmath179 and @xmath156 modes ) .
then in section [ pb ] we prove that , in the @xmath209 case , it is possible to place pointwise bounds on these fields in the outer static region , establishing in this way the non - modal stability of this region .
consider the first order perturbation fields @xmath210 where @xmath211 stands for derivative at @xmath212 for a mono parametric family of solutions of the einstein maxwell equations , as in equations ( [ lf])-([3 ] ) , the @xmath212 solution being ( [ rn])-([max ] ) , and a star denotes hodge dual @xmath213 @xmath214 being the volume form . since @xmath215 in the background , the fields @xmath208 and @xmath5 are gauge invariant @xcite and thus suitable to analyze the effects of the perturbation in the geometry and the electromagnetic field .
the obvious advantage of dealing with _
fields rather than higher rank tensor fields is that for the latter there is no entirely natural concept of being `` large '' or `` small '' in a lorentzian manifold , which we need to quantify the strength of the perturbation .
+ it follows from equations ( [ pert1 ] ) , ( [ ha ] ) and ( [ wfi ] ) , that @xmath216 depends on up to two derivatives of @xmath174 and @xmath217 on up to four derivatives of @xmath46 .
however , using repeatedly the @xmath156 leme ( [ master])-([master2 ] ) and calculating separately the @xmath179 contributions @xmath218 and @xmath219 , we can simplify considerably the resulting expressions and find , with the help of symbolic manipulation programs @xcite , that @xmath220 and @xmath221.\ ] ] the above equations allow us to prove that @xmath208 and @xmath222 contain all the gauge invariant information of a given perturbation , and also that they satisfy a coupled system of wave equations .
consider the odd solutions @xmath127 of the leme ( [ 1])-([3 ] ) around a reissner - nordstrm ( a)ds black hole background : * the map @xmath223 , \mathcal{f}_{\a \b } ] ) \to ( \f,\q)$ ] is invertible : it is possible to reconstruct @xmath124 and a representative of @xmath42 $ ] from @xmath224 . *
let @xmath225 the gauge invariant scalar fields @xmath222 and @xmath208 satisfy the system of wave equations @xmath226 \mathcal{k } = ( \j2 + 2 ) \f\\ \label{inv2 } & \left [ \nabla_{\a } \nabla^{\a } + \left ( \frac{2m}{r^3}- \frac{6q^2}{r^4}- \frac{2 \l}{3 } \right ) \right ] \left(\frac{r^3}{4q } \f \right ) = - \frac{q}{r^3 } \mathcal{k}\end{aligned}\ ] ] * expand the fields @xmath222 and @xmath208 in the orthonormal basis of spherical harmonics @xmath227 . from the the @xmath179 components of @xmath222 we can read off the @xmath206 using equation ( [ f ] ) , and from the @xmath156 components we can obtain the components of @xmath174 using this same equation : @xmath228 ( note that we are using the fact that @xmath229 is invertible in @xmath230 . ) once we have the @xmath206 and @xmath174 , we can use ( [ q ] ) to get @xmath231 from @xmath208 , from where @xmath46 can be obtained since the operator @xmath232 is invertible in @xmath230 , as it acts as @xmath233 on the @xmath81 subspace of @xmath234 .
* from ( [ f ] ) and ( [ q ] ) follows that @xmath235 then using @xmath236=0 $ ] and the @xmath156 equations ( [ master ] ) -([master2 ] ) we find that the projections @xmath217 and @xmath216 satisfy the system of equations @xmath237 \mathcal{k}_{>1 } = ( \j2 + 2 ) \f_{>1}\\ \label{inv2 } & \left [ \nabla_{\a } \nabla^{\a } + \left ( \frac{2m}{r^3}- \frac{6q^2}{r^4}- \frac{2 \l}{3 } \right ) \right ] \left(\frac{r^3}{4q } \f_{>1 } \right ) = - \frac{q}{r^3 } \mathcal{k}_{>1}\end{aligned}\ ] ] interestingly enough , the @xmath179 analogue of @xmath238 , @xmath239 together with ( c.f equations ( [ knds1])-([knds2 ] ) ) @xmath240 can also be checked to satisfy the system of equations ( [ inv1])-([inv2 ] ) ( note that the right hand side of ( [ inv1 ] ) vanishes in this case ) .
thus , equations ( [ inv1])-([inv2 ] ) follow .
note that , as a consequence of i ) in this theorem , @xmath126 in ( [ l- ] ) and ( [ l-2 ] ) can be parametrized by the set @xmath224 of solutions of the system ( [ inv1])-([inv2 ] ) or , equivalently , the set @xmath241 , \f_{\a \b}),\q([h_{\a \b } ] , \f_{\a \b}))$ ] where @xmath242 is a solution of the leme .
+ in the @xmath243 limit equations ( [ inv1 ] ) and ( [ inv2 ] ) decouple .
the the first one gives the four dimensional regge - wheeler equation for @xmath208 used in @xcite and @xcite to prove the nonmodal linear stability of the schwarzschild ds black hole whereas the second one gives the fackerrel - ipser equation for a test maxwell field on a schwarzschild ( a)ds black hole @xcite @xcite the standard way of solving the @xmath156 leme ( [ master])-([master2 ] ) is by projecting this system onto the @xmath81 subspaces and then decoupling the resulting pair of fixed @xmath81 equations by introducing two regge - wheeler fields @xcite @xcite .
this is equivalent to introducing the operator @xmath244 which is well defined and positive definite in @xmath78 , as it acts on @xmath245 as multiplication times @xmath246 and two fields @xmath247 in terms of which @xmath248 this makes the system ( [ master])-([master2 ] ) equivalent to the regge - wheeler equations , first derived in @xcite , @xmath249 where @xmath250 is a _ tortoise _ radial coordinate , defined by , @xmath251 and @xmath252 in terms of these fields , the @xmath156 piece of @xmath208 and @xmath222 are @xmath253 \j2 \phi_1 + \left [ \frac{3m+\xi}{4r}-\frac{q^2}{r^2}\right ] \j2 ( \j2 + 2 ) \phi_2\\ \label{f>1 } \f^{>1 } & = \frac{4q}{r^4 } ( 3 m + \xi ) \j2 \phi_1 + \frac{4q^2}{r^4 } ( \j2 + 2)\j2 \phi_2\end{aligned}\ ] ] up to this point , the considerations in this paper were insensitive to the value of the cosmological constant : odd perturbations can always be treated using the gauge invariant potentials @xmath46 and @xmath174 and constants @xmath206 , and theorem 1 holds irrespective of the value of @xmath22 . in the rest of this section , however , we will consider the _ evolution of initial data _ for the leme , for which we need to make a distinction between the cases @xmath254 and @xmath255 due some key differences in their causal structure .
+ in the asymptotically ads case @xmath256 , @xmath19 in ( [ f ] ) has two positive roots @xmath257 ( we will restrict for the moment to the non extremal case @xmath258 ) and the hypersurfaces they define bound three regions : i ( @xmath259 ) , ii ( @xmath260 ) and iii ( @xmath261 ) .
isometric copies of these regions are obtained by `` kruskalizing '' around the simple roots @xmath262 of @xmath38 .
this gives the maximal analytic extension depicted in figure [ adsfig ] , which extends infinitely in the vertical direction .
note that regions i and iii , where @xmath31 , are static whereas @xmath263 in region ii , which is therefore non static .
note also that the union of regions ii , ii , iii and iii fails to be globally hyperbolic due to the timelike character of the future and past null infinities @xmath264 .
this is the peculiar aspect of asymptotically anti de sitter spaces that differentiates it from asymptotically de sitter or flat spaces . in
the asymptotically ads case the dynamics of wave - like equations requires a prescription of boundary conditions at the conformal timelike boundary @xmath265 @xmath265 , which corresponds to @xmath266 .
different boundary conditions lead to different dynamics , including unstable and stable ones @xcite . for this reason ,
from now on , we restrict to the cases @xmath255 , for which the dynamic is unique and , as we will show , stable .
+ the carter - penrose diagram of ( part of ) the maximal analytic extension of the ads black hole .
the union of ii , ii , iii and iii fails to be globally hyperbolic due to the timelike character of @xmath265 , @xmath267 which is peculiar to asymptotically anti de sitter spaces . ] for @xmath268 and @xmath269 , @xmath19 has again two positive roots @xmath270 that correspond respectively to the cauchy and black hole horizons . in this case @xmath271 with @xmath272 and @xmath273 .
the outer static region , region iii in figure [ rnfig ] , corresponds to @xmath274 whereas the inner static region @xmath275 is the one defined by @xmath259 ; the singularity at @xmath276 is covered by these two horizons .
kruskalizing at @xmath277 and @xmath278 we get further copies of these regions resulting the diagram in the figure , which extends infinitely in the vertical direction .
the union of ii , ii , iii and iii is globally hyperbolic , i and i being extensions beyond the cauchy horizon at @xmath279 , which is the future boundary of the maximum cauchy development of initial data given at a complete spacelike hypersurface extending from spacelike infinity in region iii to spacelike infinity in region iii . in the extreme case @xmath280 , @xmath281 and region ii collapses . for @xmath282
the spacetime is not a black hole but an ( unstable , see @xcite and @xcite ) naked singularity . the carter - penrose diagram of ( part of ) the maximal analytic extension of the @xmath283 black hole .
the union of ii , ii , iii and iii is globally hyperbolic , its boundary at @xmath277 is a cauchy horizon . ] for @xmath284 we focus on the non extremal cases , for which @xmath19 has three simple positive roots @xmath285 which correspond to the inner , black hole and cosmological horizons respectively , and a fourth root at @xmath286 : @xmath287 @xmath288 and @xmath22 , as well as the relations among them can be found in terms of @xmath262 and @xmath289 by comparison of ( [ fr ] ) with ( [ f ] ) . as before , regions separated by the horizons are numbered in increasing number for larger @xmath290 values .
since we can kruskalize around all three horizons and large @xmath291 constant hypersurfaces are spacelike , we get the diagram in figure [ rndsfig ] , which extends infinitely in both directions .
there are a number of extremal cases corresponding to @xmath281 , @xmath292 , etc , the carter - penrose diagrams for these cases can be found in @xcite .
+ the carter - penrose diagram of ( part of ) the maximal analytic extension of a non extremal ( three different horizons ) de sitter black hole . ] in the rest of this section we will consider the stability of the outer static region iii of @xmath255 black holes , which is globally hyperbolic in these cases . in section [ cs ] we will consider the stability of the cauchy horizon separating ii from i and i , its name coming from the fact that is the boundary of the maximal future development of data given at cauchy surface of the union of ii , ii , iii and iii. + following @xcite we write ( [ rwe])-([rwp ] ) as @xmath293 where @xmath294 and @xmath295 are bounded functions on the outer static region @xmath296 for @xmath255 .
a non trivial fact , proved in section 6.2.1 of @xcite is that the @xmath297 are positive definite self adjoint operators in the space @xmath298 of square integrable functions of region @xmath296 under this particular measure .
the proof is based on a particular @xmath299deformation ( defined in @xcite ) of the @xmath300 s .
+ the proof of the following theorem is a straightforward adaptation to equation ( [ rwe ] ) of theorem 1 in @xcite for the klein gordon equation on a schwarzschild background .
it uses the self adjointness and positive character of the @xmath300 , @xmath301 and @xmath302 : [ bound ] assume @xmath303 is a solution of equation ( [ rwe ] ) on the union of regions ii , ii , iii and iii of the extended ( figure [ rnfig ] or de sitter ( figure [ rndsfig ] ) spacetimes , which has compact support on cauchy surfaces .
there exists a constant @xmath304 that depends on the datum of this field at a cauchy surface , such that @xmath305 for all points in the outer static region @xmath296 .
the argument in @xcite showing that we can restrict to fields that vanish at the bifurcation sphere together with its kruskal time derivative holds here case because the @xmath306 required isometry exchanging iii @xmath307 iii is also available in this case .
this implies that we may restrict our attention to fields in the outer static region decaying towards the bifurcation sphere as detailed in the appendix in @xcite .
+ on a @xmath308 slice of region iii , define the @xmath309 norm of a real field @xmath310 as @xmath311 note that this is _ not _ the volume element induced from the spacetime metric .
the usefulness of the above norm lies in the sobolev type inequality , equation ( 5.27 ) in @xcite , relating it with a point wise boundedness of @xmath310 on the slice @xmath312 where @xmath313 is a constant . applying this to the regge - wheeler fields @xmath303 at a fixed time @xmath308 gives @xmath314 as in the appendix in @xcite
, we will follow the strategy of proving that the @xmath309 norms on the right hand side of ( [ sobo2 ] ) can be bounded by the energies of related field configurations .
since energy is conserved for solutions of ( [ rwe2 ] ) , we get in this way a @xmath315independent upper bound of the right side of ( [ sobo2 ] ) and therefore , a global bound of @xmath316 for all @xmath317 , i.e. , of @xmath303 in the outer static region iii .
+ the conserved ( i.e. , @xmath315independent ) energy associated to equation ( [ rwe2 ] ) is @xmath318 since @xmath319 does not depend on @xmath308 , we may ( and will ) regard it as a functional on the initial datum : @xmath320 , where @xmath321 and @xmath322 : @xmath323 from ( [ aeq ] ) @xmath324 where @xmath325 is the least upper bound of @xmath326 on the outer region iii , and similarly for the other terms .
combining this with ( [ sobo2 ] ) gives @xmath327 we now use the fact that applying to a cauchy datum @xmath328 an operator that is a function of @xmath229 or @xmath300 commutes with time evolution @xcite , and also use the positive definiteness of the @xmath300 to define @xmath329 by means of the spectral theorem .
this allows us to estimate each term on the right hand side of ( [ sobo3 ] ) with the energy of field configurations related to the one with initial datum @xmath328 ( note that the first three equations below are taken verbatim from the appendix in @xcite ) , @xmath330 and to replace the right hand side of ( [ sobo3 ] ) with a time independent constant made out of the datum @xmath328 , as desired .
let @xmath222 and @xmath208 be the fields ( [ fopf ] ) associated to a solution of the leme . under the assumptions of the theorem , in the outer static region iii of a @xmath255 black hole @xmath331 where @xmath332 and @xmath333 are constants that depend on the cauchy datum @xmath334 of the perturbation .
we use that the fields @xmath335 and @xmath336 that appear in ( [ q>1 ] ) and ( [ f>1 ] ) all satisfy equation ( [ rwe2 ] ) and so , according to the theorem , are bounded by a constant .
equation ( [ invbounds ] ) follows from this observation together with equations ( [ f ] ) and ( [ q ] ) .
theorem 1.i together with the above corollary prove our notion of non - modal linear stability for the outer static region of @xmath255 black holes .
the two isometric copies of the region @xmath259 attached to the future of @xmath277 are one among infinitely many different possible extensions of the spacetime beyond @xmath277 ( although the only _ analytic _ one ) . for @xmath255 , this extension spoils the global hyperbolicity of the union of regions ii , ii , iii and iii by introducing causal curves that end in the past at the @xmath276 singularity .
regions i and i are beyond the maximal cauchy development of a spacelike surface extending from spacelike infinity ( bifurcation sphere at @xmath289 ) in iii to spacelike infinity ( bifurcation sphere at @xmath289 ) in iii in the @xmath268 ( @xmath284 ) case . in the @xmath268 case , this spacelike hypersurface is complete .
this rather disturbing feature of a solution of general relativity is considered to be _ non generic _ , in a sense yet to be made precise , and referred to as the _ strong cosmic censorship conjecture _ , first proposed by penrose almost fifty years ago @xcite . the original argument by penrose in the @xmath268 case ,
is that a small amount of radiation originating outside the black hole and coming into the non static region ii ( @xmath337 ) is gravitationally blueshifted as it propagates inwards parallel to the cauchy horizon , in a way such that the energy flux measured by an observer in free fall towards ( the right copy , see figures [ rnfig ] and [ rndsfig ] ) of region i ( @xmath259 ) diverges as @xmath338 .
this idea has proved to hold true for electro - gravitational perturbations at the linear level in @xcite , where it was shown that for a radially free falling observer with 4-velocity @xmath339 the @xmath340 piece of @xmath341 ( and therefore the complete field ) diverges for @xmath342 as the observer approaches the cauchy horizon .
although for some time it was thought that a positive cosmological constant introduces a competition of red and blueshifts effects that prevents this divergence in the case of nearly extremal ( @xmath343 ) black holes @xcite , it was later proved in @xcite that if we take into account the contributions of scattered outgoing modes the divergence occurs for any positive value of @xmath22 allowing for a three horizon structure .
further evidence of the instability of the cauchy horizon are @xcite and related works , as well as more recent models including a scalar field ( to avoid birkhoff s theorem ) , see @xcite and @xcite .
+ the regge - wheeler - zerilli potentials @xmath303 in equation ( [ hartle ] ) , whose derivative with respect to proper time is shown to diverge at the cauchy horizon in @xcite for @xmath268 and for @xmath284 in @xcite are , of course , non observable , as they are potentials for the metric and electromagnetic field perturbations , although the square of ( [ hartle ] ) contributes to the flux of energy of the perturbation .
however , the @xmath303 enter the harmonic expansion of the @xmath156 pieces of @xmath6 and @xmath5 which , according to equations ( [ f ] ) , ( [ q ] ) , ( [ wfi ] ) and ( [ pfi ] ) are @xmath344,\\ \label{qhart } \mathcal{q}_{>1 } & = \frac{q^2-mr}{r^8 } [ 2q ( r(\j2 + 2)-(3m+\xi ) ) \j2 \phi_1 + \tfrac{1}{2 } ( r ( 3m+\xi)-4q^2 ) \j2 ( \j2 + 2 ) \phi_2],\end{aligned}\ ] ] and the implications of the combined set of equations ( [ hartle])-([qhart ] ) are immediate : for a radially infalling observer crossing @xmath277 , the rate @xmath345 is clearly nonzero and finite .
for such an observer , @xmath346 commutes with the angular operators @xmath229 and @xmath302 .
this implies that both @xmath347 and @xmath348 will diverge along this geodesic as @xmath349 ( and so will diverge @xmath350 and @xmath351 ) , suggesting that the cauchy horizon is replaced with a curvature singularity .
of course , this statement needs to be taken with care since as soon as @xmath277 is approached and these quantities start to grow , the linearized equations become useless an one needs to study the evolution of the perturbation using other techniques ; but in any case the divergence at @xmath277 of the linear fields is a clear indication of strong cosmic censorship . +
the extreme case @xmath281 has been less studied , although it was recently shown that for @xmath268 ( case in which the extreme black holes corresponds to @xmath352 ) the transverse derivative @xmath353 in coordinates @xmath354 ( again , for a fixed @xmath81 component ) diverges at @xmath279 as @xmath355 along the horizon null generators @xcite .
the divergence follows from a set quantities that are shown from the ( [ rwe ] ) to be conserved along the @xmath281 horizon generators .
these are analogous to a similar set of conserved quantities for the massless scalar wave equation found in @xcite , the conservation of which was shown in @xcite to follow from a combination of newman penrose conserved quantities at null infinity @xcite and a conformal discrete isometry that exchanges the degenerate horizon and null infinity , isometry discovered in @xcite and used in @xcite to explain the symmetry of the effective potential and the consistency of the pointwise bounds for a massles scalar field in the extreme case .
+ perturbations of the extreme @xmath268 case were studied non linearly in a recent paper by reiris @xcite where it was shown that small electro - vacuum perturbations of initial data of extreme reissner - nordstrm black holes can not decay in time into an extreme kerr - newman black hole .
the evidence in @xcite is that these non - stationary solutions of the einstein - maxwell equations will settle into a sub - extremal black hole of the kerr - newman family . + as a final comment we mention that , besides the strong evidence supporting the idea that a slightly perturbed reissner - nordstrm black hole will develop a curvature singularity that cuts off the innermost region i ( @xmath356 ) , this non - unique extension beyond @xmath277 is by itself linearly unstable under electro - gravitational perturbations @xcite .
the instability , confined to this region , belongs to the even sector of the linear perturbations .
we proved that the odd sector of einstein - maxwell perturbations around a ( a)ds black hole shares with the uncharged schwarzschild black hole the property that there are physically meaningful , gauge invariant scalar fields @xmath208 and @xmath222 encoding the same information as a gauge class of a metric perturbation and satisfying a system of four dimensional wave equations which are entirely equivalent to the linearized einstein - maxwell equations
. for uncharged black holes @xmath357 the system of equations decouple , leaving the equation found in @xcite @xcite for the gravitational degrees of freedom , encoded in @xmath208 , and the fackerrel ipser equation for the maxwell degrees of freedom @xmath222 .
+ besides the significant reduction of the linearized einstein - maxwell system to scalar field equations , the resulting system of equations allow us to prove that , for generic perturbations , @xmath208 and @xmath222 are pointwise bounded in the outer static region .
this gives a strong notion of linear stability in this region , analogous to that found in @xcite and @xcite for the schwarzschild black hole .
+ if we assume that the large @xmath358 decay at fixed @xmath290 of solutions of the regge - wheeler equation in the @xmath357 case ( known as price tails , see @xcite , and @xcite for @xmath284 ) occurs also for the @xmath30 regge - wheeler fields @xmath303 in ( [ rwe2 ] ) , and use the 1 - 1 correspondence between odd perturbations and the set of @xmath359 s , together with equations ( [ f ] ) , ( [ q ] ) , ( [ wfi ] ) and ( [ pfi ] ) , we conclude that at large @xmath308 @xmath360 and @xmath361 which corresponds to a deformation within the kerr - newman ( kerr - newman de sitter ) family by adding a small amount of angular momentum .
this is consistent with the picture that the perturbed black hole settles at large times into a slowly rotating charged black hole .
+ the divergence of @xmath362 and @xmath363 for free falling radial observers as they approach the cauchy horizon from region ii , proved in the previous section , supports strong cosmic censorship in its purest form , as @xmath6 is a perturbed curvature scalar .
this result , however , has to be taken with caution , as the linear perturbation scheme becomes less reliable as linear fields grow .
this work was partially funded by grants pip 11220080102479 ( conicet - argentina ) and 30720110101569cb ( universidad
nacional de crdoba ) .
j.m.f.t . is supported by a fellowship from conicet .
99 g. dotti , `` nonmodal linear stability of the schwarzschild black hole , '' phys .
lett . * 112 * , 191101 ( 2014 ) , h. kodama and a. ishibashi , `` master equations for perturbations of generalized static black holes with charge in higher dimensions , '' prog .
phys . * 111 * , 29 ( 2004 ) [ hep - th/0308128 ] .
g. dotti , `` black hole nonmodal linear stability : the schwarzschild ( a)ds cases , '' class .
grav . * 33 * , 205005 ( 2016 ) , arxiv:1603.03749 [ gr - qc ] .
[ arxiv:1307.3340 [ gr - qc ] ] .
t. regge and j. a. wheeler , `` stability of a schwarzschild singularity , '' phys .
rev . * 108 * , 1063 ( 1957 ) .
doi:10.1103/physrev.108.1063 f. j. zerilli , `` effective potential for even parity regge - wheeler gravitational perturbation equations , '' phys . rev .
* 24 * , 737 ( 1970 ) .
doi:10.1103/physrevlett.24.737 f. j. zerilli , perturbation analysis for gravitational and electromagnetic radiation in a reissner - nordstrm geometry . " phys .
d * 9 * , 860 ( 1974 ) .
a. ishibashi and h. kodama , `` perturbations and stability of static black holes in higher dimensions , '' prog . theor .
suppl . * 189 * , 165 ( 2011 ) [ arxiv:1103.6148 [ hep - th ] ] . v. moncrief , `` odd - parity stability of a reissner - nordstrom black hole , '' phys .
d * 9 * , 2707 ( 1974 ) . v. moncrief , `` stability of reissner - nordstrom black holes , '' phys .
d * 10 * , 1057 ( 1974 ) . v. moncrief , `` gauge - invariant perturbations of reissner - nordstrom black holes , '' phys .
d * 12 * , 1526 ( 1975 ) .
r. wald , `` note on the stability of the schwarzschild metric '' , j. math .
20 , 1056 ( 1979 ) , erratum j. math .
21 , 218 ( 1980 ) .
e. chaverra , n. ortiz and o. sarbach , `` linear perturbations of self - gravitating spherically symmetric configurations , '' phys .
d * 87 * , no . 4 , 044015 ( 2013 ) [ arxiv:1209.3731 [ gr - qc ] ] .
a. ishibashi and r. m. wald , `` dynamics in nonglobally hyperbolic static space - times .
3 . anti - de sitter space - time , '' class .
* 21 * , 2981 ( 2004 ) [ hep - th/0402184 ] .
see http://grtensor.phy.queensu.ca/ e. d. fackerell and j. r. ipser , `` weak electromagnetic fields around a rotating black hole , '' phys .
d * 5 * , 2455 ( 1972 ) .
doi:10.1103/physrevd.5.2455 j. jezierski and t. smoka , `` a geometric description of maxwell field in a kerr spacetime , '' class . quant . grav .
* 33 * , no . 12 , 125035 ( 2016 ) [ arxiv:1502.00599 [ gr - qc ] ]
. b. araneda and g. dotti , in preparation .
b. s. kay and r. m. wald , `` linear stability of schwarzschild under perturbations which are nonvanishing on the bifurcation two sphere , '' class .
* 4 * , 893 ( 1987 ) .
dimock j. and kay b. s. , `` classical and quantum scattering theory for linear scalar fields on the schwarzschild metric i '' , annals of physics , * 175 * ( 1987 ) , 366 . r. h. price , `` nonspherical perturbations of relativistic gravitational collapse .
1 . scalar and gravitational perturbations , '' phys .
d * 5 * , 2419 ( 1972 ) .
doi:10.1103/physrevd.5.2419 . k. lake , `` reissner - nordstrom - de sitter metric , the third law , and cosmic censorship '' , phys .
d * 19 * , 421 ( 1979 ) .
p. r. brady , c. m. chambers , w. krivan and p. laguna , `` telling tails in the presence of a cosmological constant , '' phys .
d * 55 * , 7538 ( 1997 ) doi:10.1103/physrevd.55.7538 [ gr - qc/9611056 ]
. s. chandrasekhar , `` on algebraically special perturbations of black holes '' , proc .
a * 392 * ( 1984 ) 1 .
s. chandrasekhar , `` the mathematical theory of black holes '' , oxford university press , second edition , 1992 .
r. penrose , in _ battelle rencontres _ , edited by c. de witt and j.a .
wheeler ( w.a .
benjamin , new york , 1968 ) , p. 222 .
s. chandrasekhar and j. b. hartle , `` on crossing the cauchy horizon of a reissner - nordstrom black - hole '' , proceedings of the royal society of london . * a 384 * ( 1982 ) , 301 .
c. m. chambers , `` the cauchy horizon in black hole de sitter space - times , '' annals israel phys .
* 13 * , 33 ( 1997 ) [ gr - qc/9709025 ] .
p. r. brady , i. g. moss and r. c. myers , `` cosmic censorship : as strong as ever , '' phys .
* 80 * , 3432 ( 1998 ) doi:10.1103/physrevlett.80.3432 [ gr - qc/9801032 ] .
g. dotti , r. gleiser and j. pullin , `` instability of charged and rotating naked singularities , '' phys . lett .
b * 644 * , 289 ( 2007 ) doi:10.1016/j.physletb.2006.12.004 [ gr - qc/0607052 ] . g. dotti and r. j. gleiser , `` gravitational instability of the inner static region of a reissner - nordstrom black hole , '' class .
* 27 * , 185007 ( 2010 ) doi:10.1088/0264 - 9381/27/18/185007 [ arxiv:1001.0152 [ gr - qc ] ] .
e. poisson and w. israel , `` internal structure of black holes , '' phys .
d * 41 * , 1796 ( 1990 ) .
doi:10.1103/physrevd.41.1796 l. m. burko , `` black hole singularities : a new critical phenomenon , '' phys . rev .
lett . * 90 * , 121101 ( 2003 ) erratum : [ phys .
* 90 * , 249902 ( 2003 ) ] doi:10.1103/physrevlett.90.121101 [ gr - qc/0209084 ] .
h. maeda , t. torii and t. harada , `` novel cauchy - horizon instability , '' phys .
d * 71 * , 064015 ( 2005 ) doi:10.1103/physrevd.71.064015 [ gr - qc/0501042 ] .
j. lucietti , k. murata , h. s. reall and n. tanahashi , `` on the horizon instability of an extreme reissner - nordstrm black hole , '' jhep * 1303 * , 035 ( 2013 ) doi:10.1007/jhep03(2013)035 [ arxiv:1212.2557 [ gr - qc ] ] . s. aretakis , `` horizon instability of extremal black holes , '' adv .
theor . math .
* 19 * , 507 ( 2015 ) doi:10.4310/atmp.2015.v19.n3.a1 [ arxiv:1206.6598 [ gr - qc ] ] .
p. bizon and h. friedrich , `` a remark about wave equations on the extreme reissner - nordstrm black hole exterior , '' class .
quant . grav .
* 30 * , 065001 ( 2013 ) doi:10.1088/0264 - 9381/30/6/065001 [ arxiv:1212.0729 [ gr - qc ] ] .
e. t. newman and r. penrose , `` new conservation laws for zero rest - mass fields in asymptotically flat spacetime '' , proc .
london , ser . a 305 , 175204 ( 1968 ) .
couch w. and torrence r. , `` conformal invariance under spatial inversion of extreme reissner
nordstrm black holes '' , gen .
* 16 * ( 1984 ) , 789 .
s. dain and g. dotti , `` the wave equation on the extreme reissner - nordstrm black hole , '' class .
grav . * 30 * , 055011 ( 2013 ) doi:10.1088/0264 - 9381/30/5/055011 [ arxiv:1209.0213 [ gr - qc ] ] .
m. reiris , `` on perturbations of extreme kerr
newman black holes and their evolution , '' annales henri poincare * 16 * , no . 7 , 1551 ( 2015 ) . | following a program on black hole nonmodal linear stability initiated in phys .
rev .
lett .
* 112 * ( 2014 ) 191101 , we study odd linear perturbations of the einstein - maxwell equations around a reissner - nordstrm ( a)ds black hole .
we show that all the gauge invariant information in the metric and maxwell field perturbations is encoded in the spacetime scalars @xmath0 and @xmath1 , where @xmath2 is the weyl tensor , @xmath3 the maxwell field , a star denotes hodge dual and @xmath4 means first order variation , and that the linearized einstein - maxwell equations are equivalent to a coupled system of wave equations for @xmath5 and @xmath6 . for nonnegative cosmological constant we prove that @xmath5 and @xmath6 are pointwise bounded on the outer static region .
the fields are shown to diverge as the cauchy horizon is approached from the inner dynamical region , providing evidence supporting strong cosmic censorship . in
the asymptotically ads case the dynamics depends on the boundary condition at the conformal timelike boundary and there are instabilities if robin boundary conditions are chosen . |
the coexistence of solid and liquid phases in atomic clusters has been of experimental @xcite and theoretical @xcite interest for some time . in large clusters , as in bulk materials , static coexistence is expected theoretically@xcite and is readily observed in both experiment@xcite and simulation@xcite . however in sufficiently small clusters , simulations find that coexistence is dynamic i.e. clusters fluctuate in time between fully solid and fully liquid states @xcite .
small clusters avoid static coexistence because of the prohibitive cost of forming an interface .
this causes an s - bend in the microcanonical caloric curve @xcite and the corresponding negative heat capacities which been observed in small sodium clusters @xcite .
however , it is currently not known at what size static coexistence gives way to dynamic coexistence .
the time - scale on which dynamic coexistence occurs will be governed by the height of the energy barrier separating the solid and liquid states of the cluster @xcite . however ,
if a coexisting state exists , then this coexisting state should also be dynamically accessible .
this is not seen in at least some small clusters @xcite .
evidently , in sufficiently small clusters coexistence is unstable . at
what sizes do states of static coexistence occur ?
to determine when phase coexistence will be stable , one needs to know the relative cost of forming an interface between phases in a given cluster . a common approach to studying coexistence in clusters is to apply critical droplet theory in the capillarity approximation ( i.e. where cluster material parameters , especially surface energies , are approximated by their bulk values ) .
for instance reiss et al @xcite used the capillarity approximation to develop a critical droplet theory of melting in the canonical ensemble , with the ( spherically - symmetric ) solid fraction of a cluster acting as an order parameter .
nielsen et al @xcite developed a similar critical droplet theory in the microcanonical ensemble .
they demonstrated that a stable static coexisting state can exist in this theory near the melting point , and validated this with constant - energy simulations of a large copper cluster .
cleveland et al @xcite developed an aspherical model which dealt with static coexistence where the liquid only partially wets the solid .
thus , in principle , critical droplet theory can be used to discuss coexistence in a variety of circumstances . to study coexistence
we will work in the microcanonical ensemble as constant temperature ensembles tend to suppress the inherent inhomogeneity in coexisting systems@xcite .
constant energy conditions are realised in isolated systems with poor thermal coupling to their environment , such as clusters in an inert gas atmosphere
. consequently we will use microcanonical critical droplet ( mcd ) theory @xcite to examine the stability of static solid - liquid coexistence in small metal clusters .
we will then use constant - energy molecular dynamics simulations to test the predictions of the theory .
here we use microcanonical critical droplet ( mcd ) theory @xcite to examine the stability of static coexistence .
the spherical coexisting solid - liquid cluster of radius @xmath2 , is considered to consist of a solid core , of radius @xmath3 , and a liquid outer shell .
the entropy ( per atom ) , @xmath4 , of a coexisting solid - liquid cluster of solid fraction @xmath5 at energy ( per atom ) @xmath6 is approximated by @xmath7 where @xmath8 and @xmath9 . here @xmath10 is the bulk melting temperature , @xmath11 is the heat capacity and quantities with a superscript @xmath11 indicate values at the bulk melting point .
the solid energy , @xmath12 , and liquid energy , @xmath13 , are related to the total energy , @xmath6 , for a cluster of radius @xmath2 as follows : @xmath14 where @xmath15 ( @xmath16 is the solid - vapour interfacial energy density , @xmath17 is the solid - liquid interfacial energy density and @xmath18 is the liquid - vapour interfacial energy density ) , @xmath19 is the bulk material density and @xmath20 is a correlation length which characterizes the short - range interaction between the solid - liquid and liquid - vapour interface @xcite .
the inclusion of this final term in equation ( [ energy ] ) ensures that @xmath21 as @xmath22 .
we now show that mcd theory predicts that there is a critical cluster radius below which coexistence becomes unstable . for a given total energy per atom @xmath6 ,
we can regard @xmath13 as a function of @xmath12 and @xmath23 : @xmath24 .
we may then extremise the entropy ( given by ( [ entropy ] ) ) of the mixed cluster with respect to @xmath12 and @xmath23 at fixed @xmath6 .
it is then straightforward show that the entropy is extremised when the following conditions hold : 1 ) @xmath25 , which is the normal condition for thermal equilibrium , and 2 ) @xmath23 must satisfy the following equation @xcite : @xmath26 where l = @xmath27 is the latent heat of fusion per atom . for sufficiently negative values of @xmath6 , there are two positive solutions to ( [ quartic ] ) @xmath28 .
there is always a range of energies where @xmath29 , as when @xmath30 , @xmath31 .
in fact @xmath32 represents a local minima in the entropy , corresponding to a barrier separating the solid and liquid phases .
similarly , for values of @xmath6 where @xmath33 exists and is less than one , @xmath33 is a local maxima in the entropy , corresponding to a coexisting state ( note that @xmath34 exists only when @xmath35 , where @xmath0 is given by : @xmath36 where @xmath37 and @xmath38 . that is
if @xmath35 , there is a range of energies where a locally stable coexisting state exists . at such an energy e , the coexisting state , with solid fraction @xmath33 ,
is separated from the liquid state by an energy barrier at @xmath32 .
however , the coexisting state with solid fraction @xmath33 is not necessarily globally stable . if the coexisting state exists at an energy @xmath6 , then @xmath39 .
however , there exist a range of energies where @xmath40 only if @xmath1 , where @xmath41 thus for clusters with @xmath42 , at energies where the coexisting state exists , the state is always metastable ( @xmath43 ) . for clusters with @xmath44 ,
no coexisting state exists at any energy .
we note that some of these parameters may be ill - defined for small clusters .
nonetheless the values of @xmath0 and @xmath45 may be calculated using the capillarity approximation i.e. using bulk values ( if available ) for the material parameters . for lead
@xcite we find that @xmath46 nm ( or approximately 1 atom ) and @xmath47 nm ( or approximately 100 atoms ) .
this suggests that coexistence ought to be stable down to quite small cluster sizes , although it is well - known that the melting behaviour of sub-100 atom clusters is quite erratic due to electron - shell effects @xcite that are certainly not captured in the mcd model .
using molecular dynamics simulations of pb clusters we have found that coexistence becomes unstable at approximately 1000-atoms .
we simulated the caloric curves for three clusters : a 2073-atom icosahedron , a 1427-atom icosahedron and a 931-atom icosahedron .
here we use surface - reconstructed icosahedra which are thought to be stable in pb clusters @xcite at these sizes using a glue potential @xcite . at each energy
the cluster was equilibrated for 0.6 ns , then the kinetic energy was averaged over a further 0.6 ns .
the energy increment used was 0.2 mev / atom with energies adjusted between constant energy simulations by a uniform scaling of the kinetic energy .
figure [ caloric - md ] shows the caloric curves for the 2073 and 1427-atom clusters moving from the solid state ( on left ) to the liquid state ( on right ) . both clusters exhibit a coexisting solid - liquid state separating the solid and liquid phases .
the curves in figure [ caloric - md ] resemble those constructed in ref @xcite with their layer - by - layer model of cluster melting .
\(a ) ( b ) note the change in temperature at the transition from fully solid to coexistence and in the transition from coexistence to fully liquid . in mcd theory , while the transition from coexistence to liquid is first - order , the transition from solid to coexistence is continuous . to check whether the 1427 and 2057-atom clusters are superheated when they undergo the transition from solid to the coexisting state
, we cooled a 1427-atom cluster from the coexistence region down until the cluster froze .
figure [ meltfreeze ] shows the reverse caloric curve ( constructed using equilibration times of 4 ns with an energy increment of -0.2 mev / atom ) .
the freezing transition takes place to well within 5 mev / atom of the transition from solid to coexistence , suggesting that the simulated caloric curve is close to the equilibrium caloric curve and that the transition at -1.715 ev / atom is a first - order transition which exhibits little rounding due to finite - size effects .
the origin of this discontinuity at this transition may correspond to the creation of a minimum " volume of liquid and as such the discontinuity is likely to disappear as @xmath48 . to characterise the coexisting state , we follow cleveland et al @xcite in using the bimodality of the distribution of diffusion coefficients to distinguish solid and liquid atoms .
a sequence of snapshots at different total energies of the 1427-atom cluster is shown in figure [ mixed ] .
the total energy of the atoms in each snapshot decreases from top left to top right to bottom left to bottom right .
each atom in the snapshot has been coloured either white or grey depending on the whether the diffusion coefficient of that atom ( calculated over a carefully chosen timescale ) is low mobility ( solid ) or high mobility ( liquid ) .
the snapshots reveal that the molten lead does not fully wet the solid .
they also reveal that the liquid fraction of the coexisting cluster increases as the total energy increases .
note that the melt does not wet the solid as assumed in equation [ energy ] whereas in bulk lead , the melt wets the solid ( since in bulk lead @xmath49 ) .
this is a clear indication that the capillarity approximation has broken down at these cluster sizes .
figure [ caloric-931 ] shows the caloric curve for a 931-atom cluster constructed in the same way as those for the larger clusters in figure [ caloric - md ] .
note that the curve does not show a stable solid - liquid coexisting state but that large fluctuations in temperature occur near the melting transition in the 931-atom caloric curve . using the same method for identifying solid and liquid regions as was used in figure [ mixed ]
, one can verify that the fluctuations in temperature are correlated with the appearance of a liquid region in the otherwise solid cluster .
this is due to the dynamic coexistence between the solid state and a coexisting state . which state is globally stable ?
we have resolved this with a very long duration run near the melting transition .
figure [ trace-931 ] shows a history of the temperature ( as determined using the method above ) for a 931-atom cluster at a total energy of @xmath50 ev / atom over approximately 15 ns .
the final state of the cluster is liquid ( achieved after approximately 8 ns ) , which suggests that the fluctuations ( between 0 and 8 ns in figure [ trace-931 ] and between @xmath51 and @xmath52 ev / atom in figure [ caloric-931 ] ) are non - equilibrium precursors to melting rather than true dynamic coexistence .
mcd theory certainly suggests that at sizes @xmath53 the coexisting state is metastable with respect to the liquid rather than the solid .
however , the caloric curve in figure [ caloric-931 ] suggests that the coexisting state is metastable with respect to both the solid and liquid states , as the cluster appears to spend more time in the solid state .
nonetheless , we conclude that there is no stable coexisting state for the 931-atom cluster at any energy . in ref @xcite
, we constructed a less detailed caloric curve for 1130-atom lead clusters by coalescence .
this curve also showed no evidence for stable coexistence .
the molecular dynamics simulations and mcd theory in the capillary approximation do not agree on the size at which coexistence becomes metastable in pb clusters . however , the snapshots in figure [ mixed ] strongly suggest that the capillarity approximation has broken down at these cluster sizes ( at least when it comes to reproducing the behaviour of the simulated clusters )
. this should not be surprising as atomic clusters are interesting precisely because their properties depend strongly on size .
however , if we are to take mcd theory beyond the capillarity approximation , we would require a detailed understanding of how surface energies , and other quantities such as latent heat , depend on surface curvature and/or size .
it is not simply a matter of inserting more accurate material properties into ( [ r2 ] ) .
it is also important to recognise that the spherically symmetric geometry assumed in equation ( [ energy ] ) is inappropriate for situations involving partial wetting where @xmath54 . in ref
@xcite , a critical droplet model was developed for the partial wetting case , although it was applied in the canonical ensemble .
evidently a generalisation of mcd theory to the partially wetting case would also be desirable
. similarly it would be interesting to determine by simulation and experiment the size at which @xmath55 in pb clusters i.e. the size below which complete wetting gives way to partial wetting .
partial and complete wetting has been observed recently in isolated metal alloy clusters using transmission electron microscopy @xcite .
finally , in the simulations the transition from solid to solid - liquid coexistence was discontinuous .
in mcd theory this transition is continous as the energy of the coexisting cluster , as given by equation ( [ energy ] ) , approaches the energy of an entirely solid cluster as @xmath22 .
this limit is an artificial feature of equation ( [ energy ] ) ; if this constraint were to be relaxed so that the energy of the coexisting cluster approached some other value as @xmath22 ( for instance , @xmath56 then the transition from solid to solid - liquid would be discontinuous . provided the energy of the solid cluster and that of the limiting solid - liquid cluster differed only by surface energy , this discontinuity would disappear as @xmath48 .
in summary , mcd theory predicts that the static coexistence of solid and liquid phases can only occur in clusters above a certain size given by equation ( [ r2 ] ) . below this critical size coexistence
is metastable or unstable .
molecular dynamics simulations of pb clusters using a glue potential appear to corroborate this scenario , although the transition occurs at a larger size , @xmath57 , ( where @xmath58 ) than that predicted by mcd theory using bulk material parameters in the capillarity approximation .
indeed this discrepancy between theory and simulation is probably due to the breakdown of this approximation at small cluster sizes .
we note that pb has an unusually large value of the prefactor @xmath59 that appears in equation ( [ r2 ] ) due to a relatively small latent heat of fusion .
other metals with small latent heats ( which are correlated with low melting points ) , such as hg or the alkali metals , will also have large values of @xmath60 .
it may be that in such metals coexistence is forbidden at quite large sizes . in most metals , however , with relatively large latent heats of fusion , we might expect static coexistence to remain stable down to rather small sizes . | microcanonical critical droplet theory and molecular dynamics simulations are used to examine static coexistence between solid and liquid phases in nanoscale lead clusters .
it is shown that the theory predicts the existence of a metastable coexisting state above a critical cluster radius @xmath0 , with this state becoming stable for clusters of radius @xmath1 .
molecular dynamics simulations of lead clusters confirm the existence of stable coexisting states in 1427-atom and 2057-atom clusters but find no stable coexisting state in a 931-atom cluster . |
one of the most exciting products of the recent synthesis of superstring theory and cosmology has been that of cosmic superstrings @xcite . like their classical cousins
@xcite , cosmic superstrings were briefly considered then discarded @xcite , but for theoretical reasons rather than observational ones .
subsequent ( and largely non - perturbative ) analysis @xcite has shown that cosmic superstrings have been found to naturally arise in models of brane inflation @xcite , a string theory realization of inflationary cosmology theory .
cosmic strings and superstrings can produce a variety of astrophysical signatures including gravitational waves @xcite , ultra high energy cosmic rays @xcite , and gamma ray bursts @xcite .
very recent work has revealed a number of exciting new possibilities such as radio bursts from strings @xcite , effects on the cosmic 21 cm power spectrum @xcite , magnetogenesis @xcite , effects on the cmb at small angular scales @xcite , cmb polarization @xcite , microlensing from strings @xcite , strong lensing @xcite , and weak lensing @xcite .
prior work on gravitational radiation has focused on two processes : cusps ( whereby a segment of the string momentarily moves at the speed of light ) and kinks ( formed after two cosmic strings collide and reconnect ) @xcite . here
we study a third source : the radiation emitted from the reconnection process itself .
this is possible for cosmic superstrings because string theory allows us to explicitly construct the reconnection process and compute detailed interaction properties .
we will do this for the bosonic string but the presence of fermions in the superstring should not change the conclusions . in the first half of this article we compute the gravitational waveform resulting from fundamental cosmic superstring reconnection . in the second half we consider the likelihood of detection of this signal with current and future experiments .
+ the theory of cosmic superstring reconnection was developed in @xcite @xcite @xcite .
consider two long , straight wound bosonic strings on a 2d torus of size @xmath4 and skew angle @xmath5 as illustrated in figure 1 . in terms of the string tension @xmath6 the momenta are taken to be @xmath7^{1/2 } ( 1,0,0,0 , { \bf 0 } ) , \hspace{0.5 in } l_1 = l ( 0,1,0,0 , { \bf 0}),\ ] ] @xmath8^{1/2 } [ 1-v^2]^{-1/2 } ( 1,0,0,v , { \bf 0 } ) , \hspace{0.5 in } l_2 = l ( 0,\cos \theta,\sin \theta,0 , { \bf 0 } ) .\ ] ] these satisfy the tachyonic mass - shell conditions @xmath9 the relevant vertex operators are : ( @xmath10 ) @xmath11 where the volume @xmath12 is the product of the the transverse volume and the 2d torus ( methods to calculate @xmath13 can be found in @xcite ) . here
@xmath14 refer to the ( anti)holomorphic components of @xmath15 , so that @xmath16 .
these will scatter into some kinked configuration and the amplitude for all such processes can be summed . in the large - winding limit the probability of reconnection
is found to be @xmath17 it is also likely that there will be radiation emitted during this reconnection .
this can be included by using the reconnection process as a classical background trajectory @xmath18 upon which the radiation vertex operator must be integrated over @xcite .
then the probability to emit a radiated state of definite momentum @xmath19 will be the probability of reconnection times a factor depending on the radiated particle under consideration , @xmath20 \right|^2 .\ ] ] one might also consider the possibility that the strings will pass through each other without reconnecting but still emit radiation in the process .
this is a valid possibility but one in which the amplitude is suppressed by a factor of @xmath21 due to three final strings instead of two ( recall that a reconnected kinked string is considered a single string , whereas this process will produce two unconnected strings with accompanying coupling factor ) .
so to leading order this process can be neglected .
the reconnection trajectory was computed in @xcite by considering the vertex operator of the kinked string .
this is done by examining the operator product expansion of the straight strings , @xmath22 the taylor expansion of the exponential shows the vertex operators of the infinite tower of the produced states , which will appear kinked due to their large oscillator excitation number @xmath23 : @xmath24 since @xmath25 the result will be identical for the right - moving oscillators and so @xmath26 . now consider the state corresponding to the sum of these vertex operators ( [ kinkvertex ] ) : @xmath27 the expectation value of the position operator in the reconnection amplitude is then easily evaluated : @xmath28 \ + \ ( l \rightarrow r , z \rightarrow { \bar z } ) .
\end{aligned}\ ] ] while this is a technically correct answer , it is unsatisfactory in that it appears to treat string 1 different from string 2 even though of course these are unphysical labels .
the reason is that the placement of vertex operators has treated them in a non - symmetric fashion , expanding @xmath29 around the location of @xmath30 . to remedy this we simply redo the previous calculation but expand both vertex operators around the midpoint of their separation , @xmath31 a good check that this is a physically equivalent solution
( i.e. obtainable via a conformal transformation @xmath32 ) is that for the massless kink state ( @xmath33 ) the polarization vector @xmath34 has merely been replaced with @xmath35 , which is a difference proportional to the wavevector @xmath36 and thus pure gauge . now combining this new kink vertex operator with symmetric relocations @xmath37 for the incoming straight - string vertex operators , the transformed trajectory is @xmath38 \\ & - & i { \alpha ' \over 2 } p_{2l } \left [ \ln \left(z+ { 1 \over 2 } \right ) + \sum _ { n=1}^n \left ( \frac{1}{n } - \frac{1}{n+1 } \right ) \left ( - \frac{z}{2 } \right ) ^n \right ] \ + \ ( l \rightarrow r , z \rightarrow { \bar z } )
. \end{aligned}\ ] ] as hoped , this is symmetric under @xmath39 by exchanging @xmath40 ( up to subtleties involving branch cuts ) .
this simplifies in the cosmic string limit ( @xmath41 ) by performing the summation , @xmath42 we will neglect the second term since this is a good approximation for most of the parameter space , and the trajectory simplifies tremendously : @xmath43 however , the trajectory ( [ xcltemp ] ) as written is complex , due to the analytic continuation required by string theory ; we now need to transform this back into a trajectory in real space .
this is easily done by recalling that vertex operators are inserted into the path integral as source terms via @xmath44 this produces a trajectory @xmath45 we see the winding term is real , and that moving around the vertex operator produces @xmath46 as desired .
the momenta term , however , is purely imaginary ; this can be remedied by rotating @xmath47 , allowing us to interpret the ( logarithmic ) distance away from the vertex operator as displacement in time .
thus the trajectory for the reconnection process in real target space is @xmath48 \hspace{1 in } \\
\nonumber & & \hspace{1 in } + \frac{1}{4 \pi i } \left [ l_{1}^\mu \ln \left ( \frac{2 z-1}{2 { \bar z}-1 } \cdot \frac{2 - { \bar z}}{2 - z } \right ) + l_{2}^\mu \ln \left ( \frac{2 z+1}{2 { \bar z}+1 } \cdot \frac{2 + { \bar z}}{2 + z } \right ) \right ]
. \end{aligned}\ ] ] this reconnection process is shown in figure 2 , from the worldsheet and spacetime viewpoints .
the infinite past ( containing the two incoming asymptotic states of simple wound strings ) are located at @xmath49 , the infinite future ( containing the single outgoing asymptotic state of a sharply kinked reconnected string ) is located at @xmath50 , and the entire unit circle @xmath51 maps to time @xmath52 . as expected , string theory has smoothed out what is classically a very violent and singular process .
+ + + + + + now let us compute the radiation emitted from such a reconnection process .
the simplest and most physically relevant vertex operator is that for a graviton , @xmath53 where we have factored the left- and right - moving components .
our task now is to evaluate the amplitude @xmath54\ ] ] given the trajectory ( [ xcl ] ) .
as explained in more detail in section 3.1 , this problem is tantamount to determining the cosmic string s energy - momentum tensor @xmath55 damour and vilenkin investigated this quantity for classical cosmic strings @xcite , finding that it is largest when both the left- and right - moving components of the exponential have either a saddle point or discontinuity at the same location .
the signal strength can then be conveniently ranked according to the phase behavior : 1 . *
cusp * : double saddle - point .
* kink * : one saddle - point , one discontinuity .
* reconnection * : double discontinuity . since the contribution of radiation emitted from reconnection would appear to be a _
sub_-sub - leading correction , it is not surprising that henceforth no effort has been made to explicitly calculate this .
the reason we believe this is now a worthwhile endeavor is because in superstring theory the probability of reconnection @xmath56 can be greatly suppressed , @xmath57 . as discussed in more detail in the next section , this then produces a larger number of strings @xmath58 , the number of reconnection attempts now scaling as @xmath59 , and the number of successful reconnections scaling as @xmath60 .
thus , as is the case for bursts from cusps , the lower the probability of reconnection , the larger the gravitational wave signal . let us make a naive attempt to evaluate the amplitude ( [ tmunu ] ) .
in examining @xmath61 in ( [ xcl ] ) , there do not appear to be any kink - like discontinuities , but there do appear to be cusps as indicated by the presence of extrema .
focusing now on just the holomorphic part of @xmath61 , the cusp condition is @xmath62 rewriting this as a quadratic equation , one sees there are two solutions , @xmath63 one solution will be @xmath64 and the other @xmath65 . in the large - winding limit @xmath66 , away from these extrema
the phase will oscillate wildly and so we attempt to utilize a saddle - point approximation , @xmath67 \times ( l \rightarrow r , z \rightarrow { \bar z } ) \hspace{0.3 in } { \rm ( naive ) } .\ ] ] this could not be possibly be correct , as this @xmath68 spectrum indicative of simultaneous cusps does not even remotely reproduce the @xmath69 spectrum due to simultaneous kinks , which we must obtain in the classical limit .
where did we go wrong ?
the answer lies in the fact that our saddle - point evaluation was done at the edge of moduli space @xmath65 , where such an approximation breaks down @xcite .
to evaluate the amplitude correctly , we follow kawai - lewellen - tye @xcite and write the exponential part of the amplitude explicitly : @xmath70 denoting @xmath71 , the integrand is an analytic function of @xmath72 with eight branch points : @xmath73 .
the contour over @xmath72 stretches from @xmath74 to @xmath75 along the real axis , so we can then rotate it counter - clockwise to now run from @xmath74 to @xmath75 along the imaginary axis . if we then define the _ independent _ variables @xmath76 we can rewrite the integral in terms of these variables , paying special attention to the phases , @xmath77 k \cdot p_{1l } + \left [ \arg ( z_l + \frac{1}{2 } ) - \arg(2+z_l ) \right ] k \cdot p_{2l } - \left [ \arg ( z_r- \frac{1}{2 } ) - \arg(2-z_r ) \right ] k \cdot p_{1r } - \left [ \arg ( z_r + \frac{1}{2 } ) - \arg(2+z_r ) \right ] k \cdot p_{2r } \ } } .\end{aligned}\ ] ] initially lie on the real axis ( above ) and are chosen to circumnavigate the poles so as to produce the correct phases in the amplitude .
these can then be deformed into something more convenient for explicit evaluation , such as the complex unit circle with negligible contribution from infinity ( below).,title="fig:",width=249 ] initially lie on the real axis ( above ) and are chosen to circumnavigate the poles so as to produce the correct phases in the amplitude .
these can then be deformed into something more convenient for explicit evaluation , such as the complex unit circle with negligible contribution from infinity ( below).,title="fig:",width=249 ] initially lie on the real axis ( above ) and are chosen to circumnavigate the poles so as to produce the correct phases in the amplitude .
these can then be deformed into something more convenient for explicit evaluation , such as the complex unit circle with negligible contribution from infinity ( below).,title="fig:",width=249 ] initially lie on the real axis ( above ) and are chosen to circumnavigate the poles so as to produce the correct phases in the amplitude .
these can then be deformed into something more convenient for explicit evaluation , such as the complex unit circle with negligible contribution from infinity ( below).,title="fig:",width=249 ] + approximation ( blue).,width=288 ] + an examination of these phases produces the following summary : [ cols="^,^",options="header " , ] these phases can be produced by choosing the contours shown in figure 3 .
once these choices have been made , of course , we are free to deform them into something more convenient for analysis .
we choose to deform the contour into the complex unit circle , which we recall corresponds to zero time as shown in figure 2 .
note that the trajectory at this moment resembles a smoothed - out form of the classical reconnection picture .
the idea of transforming an amplitude at a given point on the string over all time into over the entire string at a single moment in time is not without precedent in superstring theory @xcite .
now that we have completely factored the left- and right moving components to ( [ tmunu ] ) , we are only concerned with the evaluation of the integral over the angle @xmath78 , @xmath79 defining the scale - invariant quantities @xmath80 , we can expand ( [ il ] ) in powers of @xmath4 near the points @xmath81 at which the trajectory is very nearly linear : @xmath82 e^{-ik \cdot \left [ \psi \partial_\phi { \hat x}_l(1 ) + \frac{\psi^2}{2l } \partial_\phi^2 { \hat x}_l(1 ) + \cdots \right ] /2 } \\ \nonumber & + & \int_{l \pi /2 } ^{-l \pi /2 } d \psi \ \left [ \partial_\phi { \hat x}_l^\mu(-1 ) + \frac{\psi}{l } \partial^2_\phi { \hat x}^\mu(-1 ) + \cdots \right ] e^{-ik \cdot \left [ \psi \partial_\phi { \hat x}_l(-1 ) + \frac{\psi^2}{2l } \partial_\phi^2 { \hat x}_l(-1 ) + \cdots \right ] /2 } \\
\label{il2 } & \approx & 2i \left [ \frac { \partial_\phi { \hat x}_l^\mu(1)}{k \cdot { \partial_\phi { \hat x}_l(1 ) } } - \frac { \partial_\phi { \hat x}_l^\mu(-1)}{k \cdot { \partial_\phi { \hat x}_l(-1 ) } } \right ] \\ \nonumber & = & 2i \left [ \frac { { \hat p}_{1l}^\mu } { k \cdot { \hat p}_{1l } - \frac{9}{80 } \left [ k \cdot { \hat p}_{2l } + \frac{(k \cdot { \hat p}_{1l})^2 } { k \cdot { \hat p}_{2l } } \right ] } - \frac { { \hat p}_{2l}^\mu } { k \cdot { \hat p}_{2l } - \frac{9}{80 } \left [ k \cdot { \hat p}_{1l } + \frac{(k \cdot { \hat p}_{2l})^2 } { k \cdot { \hat p}_{1l } } \right ] } \right ] + \mathcal o \left ( \frac{1}{k^2 l } \right ) . \end{aligned}\ ] ]
thus in accordance with the classical expectation , we have obtained a spectrum which is scale - invariant in the large - winding limit , and which has virtually identical dependence upon the relative angle and velocity of the strings . in the last line
we have written it to most resemble the classical answer plus corrections ; the terms proportional to @xmath83 are the corrections from the string path integral .
of course the procedure is then repeated for the right - moving side .
we must also average over all geometric possibilities for which reconnection takes place using the probability measure @xmath84 modulated by some string distribution @xmath85 and then finally to sum over polarizations . while ( [ pmeasure ] ) itself easily evaluates to @xmath86 , averaging ( [ il2 ] ) must be done numerically .
though we have considered massless gravitational modes , the same procedure will work for any string mode by using the appropriate vertex operator in ( [ amp ] ) .
however , analysis of the decay modes for highly - excited strings suggests that radiation into massive states is suppressed and that only massless quanta are excited @xcite .
in addition to the fundamental cosmic superstrings considered here , there are also d- and @xmath87-strings produced in brane inflation which will have significantly larger tension are therefore might be capable of emitting higher - energy bursts @xcite @xcite . though the method of calculation for reconnection interactions is very different @xcite @xcite @xcite @xcite , ideally their waveform could be estimated by adapting the f - string result ( [ il2 ] ) and appropriately replacing the inverse tension as @xmath88 in the bps limit , depending on the specific brane inflation model .
these @xmath87 strings also form more sophisticated networks @xcite and thus would have distinctive burst patterns @xcite . for verification and comparison
we now compute the string trajectory of two classical nambu - goto strings during reconnection .
this is done using the geometry in figure 1 , taking the point of reconnection to be @xmath89 and the gauge conditions for the spatial coordinates @xmath90 to be @xmath91 whereas the time coordinate is defined to be @xmath92 . though there exist analytic and numerical studies of the reconnection process @xcite
@xcite @xcite , here we take the simplistic choice of instantaneous reconnection . at time
@xmath93 we can write the spatial derivative and velocity for the first string as @xmath94 @xmath95 and for string 2 as @xmath96 @xmath97 now defining @xmath98 , we can always write @xmath99 and then these tangent vectors combine into @xmath100 @xmath101 these are then utilized in the evaluation of the integral ( [ il ] ) , which is easily done near a discontinuity : @xmath102 .\end{aligned}\ ] ] to compare these to the superstring results , recall that in ( [ pl ] ) we have defined the left / right momentum vectors @xmath103 , which can then be normalized as ( substituting @xmath104 , @xmath105 ) , @xmath106 the classical result ( [ dx2 ] ) can then be directly compared to the superstring result ( [ il2 ] ) in terms of the geometry at the moment of reconnection , as shown in figure 4 .
in a network of cosmic superstrings reconnections can occur in two situations , ( 1 ) when two long strings meet , and ( 2 ) when loops are formed .
the rate at which reconnections take place in a network of superstrings , and the gravitational wave bursts that result , can be readily estimated .
if we assume the string density to be inversely proportional to the reconnection probability , at cosmic time @xmath107 the density of long strings in the scaling regime is @xcite @xmath108 where @xmath109 is the mass per unit length in strings and @xmath110 is the reconnection probability .
the total energy in long strings is @xmath111 and the total length is @xmath112 , so the total number of long strings in a hubble volume at any time is @xmath113 @xcite .
strings move relativistically so that in a time @xmath107 they travel a distance @xmath114 and each string attempts to reconnect with every other string in a hubble volume .
although the number of reconnection attempts in a time @xmath107 is @xmath115 the number of successful reconnections is @xmath116 .
so the number of long string reconnections that take place per unit space - time volume is @xmath117 reconnections also occur when loops are formed .
the scaling network converts a length of string @xmath112 into loops in a hubble time @xmath114 per hubble volume ( @xmath118 ) . if loops have a size @xmath119 at formation , the number of loops formed in a time @xmath107 is @xmath120 , so the number of reconnections per unit space - time volume is @xmath121 if @xmath122 is comparable to the hubble length then the loop contribution eq .
( [ e : nl ] ) to the burst rate is comparable to the string contribution eq .
( [ e : nr ] ) , but if the size of loops @xmath122 is determined by gravitational back - reaction then the loop contribution dominates .
the total number of reconnections per unit spacetime volume is thus , @xmath123 if @xmath124 .
the gravitational wave strain produced at the reconnection event is of order @xmath125 so that for a reconnection occurring at a redshift @xmath126 , @xmath127 the amplitude of the waveform is suppressed by an additional factor of @xmath128 due to the redshifting of the frequency @xmath129 ( see @xcite ) .
the hubble parameter today is @xmath130 km@xmath131s@xmath131mpc @xcite ) , and we have written the proper distance as @xmath132 . unlike cusps , these bursts do not point in any particular direction
. however , we do not observe bursts at arbitrarily low frequencies , rather , only down to the inverse of the scale at which the strings curve ( @xmath133 ) and the @xmath134 behaviour of the waveform is no longer valid .
using eq .
( [ e : nl2 ] ) we can write the number of bursts originating from a volume at redshift @xmath126 as @xmath135 where the factor of @xmath136 comes from the relation between the observed burst rate and the cosmic time , and the @xmath137-function cuts off events that we would not observe below frequency @xmath129 .
we also take the volume element at a redshift @xmath126 to be @xmath138 and so the rate of events from redshifts between @xmath126 and @xmath139 is @xmath140 to compute the observable rate of bursts we use the similar methods to compute sensitivity estimates as used in @xcite .
we write the strain produced by the reconnection cusp as @xmath141 with @xmath142 using the noise curve eq .
( 39 ) of @xcite we find a minimum detectable amplitude @xmath143 s for initial ligo and @xmath144 s for advanced ligo . using eq .
( [ e : amplitude1 ] ) we can find what redshifts these minimum amplitudes correspond to for a given value of @xmath145 ( @xmath146 , say ) and integrate eq .
( [ e : drdz ] ) to find the rate @xmath147 the cosmological functions that enter these expressions , @xmath148 , @xmath149 , and @xmath150 depend on cosmological parameters and need to be computed numerically as described in appendix a of @xcite , where a vanilla @xmath151-cdm model was used @xcite .
however , one can obtain fairly good analytic approximations for the cosmological functions using variations of the expressions introduced in @xcite .
in particular the following expressions , @xmath152 with @xmath153 , @xmath154 and @xmath155 agree to better than 20% with numerically computed functions at all relevant redshifts .
we will use these functions to compute the burst rates and stochastic backgrounds .
s , @xmath156 hz ) as a function of string tension for a reconnection probability @xmath157 .
the solid , dotted and dashed - dotted lines show the number of events for loop sizes given by gravitational back - reaction @xcite @xmath158 with @xmath159 , @xmath160 , and @xmath161 respectively .
we have taken @xmath162 everywhere .
the dashed line shows the burst rate for loop sizes determined by the large scale dynamics of the network .
we have taken @xmath163 as suggested by recent numerical simulations @xcite .
unfortunately , none of these models result in a signal detectable by ligo.,width=576 ] figure [ fig : bursts ] shows the number of burst events per year as a function of string tension for for advanced ligo as a function of string tension for a reconnection probability @xmath157 .
we have taken the minimum detectable amplitude to be @xmath144 s , set the frequency to @xmath156 hz , and taken @xmath162 everywhere .
the solid , dotted and dashed - dotted lines show the number of events for loop sizes given by gravitational back - reaction @xcite @xmath158 with @xmath159 , @xmath160 , and @xmath161 respectively .
the dashed line shows the burst rate for large long lived loops ( @xmath164 ) as suggested by recent numerical simulations @xcite . unfortunately
, none of these models result in a signal detectable by ligo . to compute the stochastic background of gravitational waves generated by all the reconnections occurring in the network we can use the results in @xcite for cusps .
the spectrum of gravitational waves is given by @xmath165 where @xmath166 excludes large bursts that occur infrequently and contaminate the estimate of the strain @xcite , and @xmath167 is the strain produced at frequency @xmath129 by a burst from a reconnection at redshift @xmath126 . as a function of the string tension for @xmath168 and @xmath169 .
the solid lines show the stochastic background generated at @xmath170 hz , corresponding to ligo frequencies , the dotted line shows the spectrum at @xmath171 hz , the band at which pulsar timing experiments are most sensitive .
again , unfortunately , none of these models result in a detectable signal .
although not shown here , large loops ( @xmath163 ) also do not result in a detectable signal.,width=576 ] figure [ fig : stoch ] shows three examples of the value of @xmath172 at the frequencies of various experiments as a function of the string tension for @xmath157 .
the solid lines show the stochastic background generated at @xmath170 hz , corresponding to ligo frequencies , for two values of the loop size .
the dotted line shows the spectrum at @xmath171 hz , the band at which pulsar timing experiments are most sensitive .
again , unfortunately , none of these models result in a detectable signal .
although not shown here , large loops ( @xmath163 ) do not result in a detectable signal .
we have calculated the gravitational waveform due to cosmic superstring reconnection , finding that bursts do not occur frequently enough to be seen with upcoming gravitational wave detectors , nor is the stochastic background strong enough to be detected .
this conclusion remains true even for very auspicious values of currently unknown parameters such as reconnection probability and loop sizes .
thus the most likely source of cosmic string gravitational bursts remains that resulting from cusps or kinks .
this is doubly disappointing since one very attractive feature of the spectrum is that the gravitational radiation is polarized along the directions of @xmath173 at reconnection and so there must be a consistency relation with the probability distribution @xmath174 encoded in the kinks on horizon - sized strings .
lisa , however , may be sensitive enough to detect bursts from reconnections and a detailed calculation including confusion noise from the network is underway . despite this disappointing result from gravitational bursts ,
the techniques we have developed here can immediately be generalized to other ( non - gravitational ) particles , which could be observed in complementary experiments such as gamma - ray bursts or other astrophysical phenomena .
since the suppression is only power - law and not exponential in energy , there remains the possibility of producing very high energy particles .
we would like to thank a. achcarro , j. j. blanco - pillado , t. damour , g. dvali , m. sakellariadou and a. stebbins for useful discussions , and especially j. polchinski and a. vilenkin for finding a critical mistake in the first draft of this article .
mgj is supported by the doe at fermilab and would like to thank the organizers of the strings and superstrings in observational cosmology workshop at apc - paris .
in the weak field limit the metric is assumed to be that of a perturbed minkowski space @xmath175 with @xmath176 . in the harmonic gauge , defined by @xmath177the linearised einstein equations read @xmath178 where @xmath179 is the stress - energy tensor of the source . the retarded potential solution to eq .
( [ lineq ] ) is given by @xmath180 for sources localised in space and time we can construct the stress - energy tensor from the fourier integral @xmath181 with the inverse transformation given by @xmath182 the retarded potential solution is @xmath183 we replace @xmath184 with @xmath185 in the denominator of the integrand in eq .
( [ retardedfourierint ] ) and in the exponent approximate @xmath186 .
this yields @xmath187 if we now substitute for @xmath188 using eq.([fourierinverse ] ) we get @xmath189 we can thus write the solution as the integral over the plane waves @xmath190 where @xmath191 is the polarisation tensor , and @xmath192 is the fourier transform of the stress - energy tensor .
99 j. polchinski , `` introduction to cosmic f- and d - strings , '' arxiv : hep - th/0412244 .
t. w. b. kibble , `` topology of cosmic domains and strings , '' j. phys . a * 9 * , 1387 ( 1976 ) .
e. witten , `` cosmic superstrings , '' phys .
b * 153 * , 243 ( 1985 ) .
e. j. copeland , r. c. myers and j. polchinski , `` cosmic f- and d - strings , '' jhep * 0406 * , 013 ( 2004 ) [ arxiv : hep - th/0312067 ] k. becker , m. becker and a. krause , `` heterotic cosmic strings , '' phys .
d * 74 * , 045023 ( 2006 ) [ arxiv : hep - th/0510066 ] .
n. t. jones , h. stoica and s. h. h. tye , `` brane interaction as the origin of inflation , '' jhep * 0207 * , 051 ( 2002 ) [ arxiv : hep - th/0203163 ] s. sarangi and s. h. h. tye , `` cosmic string production towards the end of brane inflation , '' phys .
b * 536 * , 185 ( 2002 ) [ arxiv : hep - th/0204074 ] g. dvali and a. vilenkin , `` formation and evolution of cosmic d - strings , '' jcap * 0403 * , 010 ( 2004 ) [ arxiv : hep - th/0312007 ] n. t. jones , h. stoica and s. h. h. tye , `` the production , spectrum and evolution of cosmic strings in brane inflation , '' phys .
b * 563 * , 6 ( 2003 ) [ arxiv : hep - th/0303269 ] t. damour and a. vilenkin , `` gravitational wave bursts from cosmic strings , '' phys .
* 85 * , 3761 ( 2000 ) [ arxiv : gr - qc/0004075 ] .
t. damour and a. vilenkin , `` gravitational wave bursts from cusps and kinks on cosmic strings , '' phys .
d * 64 * , 064008 ( 2001 ) [ arxiv : gr - qc/0104026 ] .
t. damour and a. vilenkin , `` gravitational radiation from cosmic ( super)strings : bursts , stochastic background , and observational windows , '' phys . rev .
d * 71 * , 063510 ( 2005 ) [ arxiv : hep - th/0410222 ] .
x. siemens , j. creighton , i. maor , s. ray majumder , k. cannon and j. read , `` gravitational wave bursts from cosmic ( super)strings : quantitative analysis and constraints , '' phys .
d * 73 * , 105001 ( 2006 ) [ arxiv : gr - qc/0603115 ] .
c. j. hogan , `` gravitational waves from light cosmic strings : backgrounds and bursts with large loops , '' phys .
d * 74 * , 043526 ( 2006 ) [ arxiv : astro - ph/0605567 ] .
x. siemens , v. mandic and j. creighton , `` gravitational wave stochastic background from cosmic ( super)strings , '' phys . rev .
lett . * 98 * , 111101 ( 2007 ) [ arxiv : astro - ph/0610920 ] . m. r. depies and c. j. hogan , `` stochastic gravitational wave background from light cosmic strings , '' phys .
d * 75 * , 125006 ( 2007 ) [ arxiv : astro - ph/0702335 ] .
v. berezinsky , b. hnatyk and a. vilenkin , `` gamma ray bursts from superconducting cosmic strings , '' phys .
d * 64 * , 043004 ( 2001 ) [ arxiv : astro - ph/0102366 ] .
bhattacharjee and g. sigl , `` origin and propagation of extremely high energy cosmic rays , '' phys .
rept . * 327 * , 109 ( 2000 ) [ arxiv : astro - ph/9811011 ] .
t. vachaspati , `` cosmic sparks from superconducting strings , '' arxiv:0802.0711 [ astro - ph ] .
r. khatri and b. d. wandelt , `` cosmic ( super)string constraints from 21 cm radiation , '' phys .
lett . * 100 * , 091302 ( 2008 ) [ arxiv:0801.4406 [ astro - ph ] ] .
l. pogosian , s. h. tye , i. wasserman and m. wyman , `` cosmic strings as the source of small - scale microwave background anisotropy , '' arxiv:0804.0810 [ astro - ph ] .
d. baumann _ et al .
_ , `` cmbpol mission concept study : a mission to map our origins , '' arxiv:0811.3911 [ astro - ph ] .
d. baumann _ et al .
_ , `` cmbpol mission concept study : probing inflation with cmb polarization , '' arxiv:0811.3919 [ astro - ph ] . d. f. chernoff and s. h. h. tye , `` cosmic string detection via microlensing of stars , '' arxiv:0709.1139 [ astro - ph ] .
k. kuijken , x. siemens and t. vachaspati , `` microlensing by cosmic strings , '' arxiv:0707.2971 [ astro - ph ]
. m. a. gasparini , p. marshall , t. treu , e. morganson and f. dubath , `` direct observation of cosmic strings via their strong gravitational lensing effect : i. predictions for high resolution imaging surveys , '' arxiv:0710.5544 [ astro - ph ] .
j. polchinski , `` collision of macroscopic fundamental strings , '' phys .
b * 209 * , 252 ( 1988 ) .
m. g. jackson , n. t. jones and j. polchinski , `` collisions of cosmic f- and d - strings , '' jhep * 0510 * , 013 ( 2005 ) [ arxiv : hep - th/0405229 ] .
m. g. jackson , `` interactions of cosmic superstrings , '' jhep * 0709 * , 035 ( 2007 ) [ arxiv:0706.1264 [ hep - th ] ] .
m. g. jackson , `` cosmic superstring scattering in backgrounds , '' jhep * 0609 * , 071 ( 2006 ) [ arxiv : hep - th/0608152 ] .
d. j. gross and p. f. mende , `` string theory beyond the planck scale , '' nucl .
b * 303 * , 407 ( 1988 ) .
h. kawai , d. c. lewellen and s. h. h. tye , `` a relation between tree amplitudes of closed and open strings , '' nucl .
b * 269 * , 1 ( 1986 ) .
a. hashimoto and i. r. klebanov , `` scattering of strings from d - branes , '' nucl .
* 55b * , 118 ( 1997 ) [ arxiv : hep - th/9611214 ] .
d. chialva , r. iengo and j. g. russo , `` decay of long - lived massive closed superstring states : exact results , '' jhep * 0312 * , 014 ( 2003 ) [ arxiv : hep - th/0310283 ] .
d. chialva and r. iengo , `` long lived large type ii strings : decay within compactification , '' jhep * 0407 * , 054 ( 2004 ) [ arxiv : hep - th/0406271 ] .
j. h. schwarz , `` an sl(2,z ) multiplet of type iib superstrings , '' phys .
b * 360 * , 13 ( 1995 ) [ erratum - ibid .
b * 364 * , 252 ( 1995 ) ] [ arxiv : hep - th/9508143 ] .
m. g. jackson , `` a note on cosmic ( @xmath193 ) strings , '' phys .
d * 75 * , 087301 ( 2007 ) [ arxiv : hep - th/0610059 ] .
a. hanany and k. hashimoto , `` reconnection of colliding cosmic strings , '' jhep * 0506 * , 021 ( 2005 ) [ arxiv : hep - th/0501031 ] .
k. hashimoto and d. tong , `` reconnection of non - abelian cosmic strings , '' jcap * 0509 * , 004 ( 2005 ) [ arxiv : hep - th/0506022 ] .
b. shlaer and m. wyman , `` cosmic superstring gravitational lensing phenomena : predictions for networks of ( @xmath194 ) strings , '' phys .
d * 72 * , 123504 ( 2005 ) [ arxiv : hep - th/0509177 ] .
s. h. tye , i. wasserman and m. wyman , `` scaling of multi - tension cosmic superstring networks , '' phys .
d * 71 * , 103508 ( 2005 ) [ erratum - ibid .
d * 71 * , 129906 ( 2005 ) ] [ arxiv : astro - ph/0503506 ] .
m. sakellariadou and h. stoica , `` dynamics of f / d networks : the role of bound states , '' jcap * 0808 * , 038 ( 2008 ) [ arxiv:0806.3219 [ hep - th ] ] .
r. brandenberger , h. firouzjahi , j. karouby and s. khosravi , `` gravitational radiation by cosmic strings in a junction , '' jcap * 0901 * , 008 ( 2009 ) [ arxiv:0810.4521 [ hep - th ] ] .
cosmic strings and other topological defects , " a. vilenkin , e.p.s .
shellard , cambridge university press ( 1994 ) .
a. achucarro and r. de putter , `` effective non - intercommutation of local cosmic strings at high collision speeds , '' phys .
d * 74 * , 121701 ( 2006 ) [ arxiv : hep - th/0605084 ] .
l. m. a. bettencourt and t. w. b. kibble , `` nonintercommuting configurations in the collisions of type i @xmath195 cosmic strings , '' phys . lett .
b * 332 * , 297 ( 1994 ) [ arxiv : hep - ph/9405221 ] .
e. p. s. shellard , `` cosmic string interactions , '' nucl .
b * 283 * , 624 ( 1987 ) ; r.a .
matzner , `` interaction of @xmath195 cosmic strings : numerical intercommutation , '' comput .
* 2 * , 51 ( 1989 ) ; k. j. m. moriarty , e. myers and c. rebbi , j. comput .
* 88 * , 467 ( 1990 ) .
m. sakellariadou , `` a note on the evolution of cosmic string / superstring networks , '' jcap * 0504 * , 003 ( 2005 ) [ arxiv : hep - th/0410234 ] .
s. eidelman _ et al . _
[ particle data group ] , `` review of particle physics , '' phys .
b * 592 * ( 2004 ) 1 .
o. lahav and a. r. liddle , `` the cosmological parameters 2005 , '' arxiv : astro - ph/0601168 .
x. siemens and k. d. olum , `` gravitational radiation and the small - scale structure of cosmic strings , '' nucl . phys .
b * 611 * , 125 ( 2001 ) [ erratum - ibid .
b * 645 * , 367 ( 2002 ) ] [ arxiv : gr - qc/0104085 ] .
x. siemens , k. d. olum and a. vilenkin , `` on the size of the smallest scales in cosmic string networks , '' phys .
d * 66 * , 043501 ( 2002 ) [ arxiv : gr - qc/0203006 ] .
f. dubath , j. polchinski and j. v. rocha , `` cosmic string loops , large and small , '' phys .
d * 77 * , 123528 ( 2008 ) [ arxiv:0711.0994 [ astro - ph ] ] . c. ringeval , m. sakellariadou and f. bouchet , `` cosmological evolution of cosmic string loops , '' jcap * 0702 * , 023 ( 2007 ) [ arxiv : astro - ph/0511646 ] .
j. polchinski and j. v. rocha , `` cosmic string structure at the gravitational radiation scale , '' phys .
d * 75 * , 123503 ( 2007 ) [ arxiv : gr - qc/0702055 ] .
j. polchinski and j. v. rocha , `` analytic study of small scale structure on cosmic strings , '' phys .
d * 74 * , 083504 ( 2006 ) [ arxiv : hep - ph/0606205 ] .
k. d. olum and v. vanchurin , `` cosmic string loops in the expanding universe , '' phys .
d * 75 * , 063521 ( 2007 ) [ arxiv : astro - ph/0610419 ] . | we compute the gravitational waveform produced by cosmic superstring reconnections .
this is done by first constructing the superstring reconnection trajectory , which closely resembles that of classical , instantaneous reconnection but with the singularities smoothed out due to the string path integral .
we then evaluate the graviton vertex operator in this background to obtain the burst amplitude .
the result is compared to the detection threshold for current and future gravitational wave detectors , finding that neither bursts nor the stochastic background would be detectable by advanced ligo .
this disappointing but anticipated conclusion holds even for the most optimistic values of the reconnection probability and loop sizes . gravitational wave bursts from cosmic superstring reconnections * mark g. jackson@xmath0 and xavier siemens@xmath1 * @xmath2particle astrophysics center and theory group fermi national accelerator laboratory batavia , illinois 60510 , usa @xmath3lorentz institute for theoretical physics leiden 2333ca , the netherlands [email protected]_ @xmath1center for gravitation and cosmology department of physics university of wisconsin - milwaukee p.o .
box 413 wisconsin , 53201 , usa [email protected]_ |
the discovery of the medium effects on jets produced at the relativistic heavy - ion collider ( rhic ) has contributed greatly to the understanding of the physics underlying hard partons traversing a hot and dense quark - gluon system@xcite .
subsequently , a large number of experimental investigations have revealed details of the properties of jets , both on the near side and on the away side , in heavy - ion collisions ( hic ) @xcite-@xcite .
particle correlations within jets are shown to have distinctive properties in the azimuthal angles @xmath1 and in pseudorapidities @xmath0 of the associated particles .
of particular noteworthiness is the peak in @xmath8 between the @xmath0 values of the trigger and the associated particles on the near side : it sits above a flat plateau or a pedestal @xcite .
no pedestal is found in @xmath9 , although a peak in that variable is prominent .
it is our aim in this paper to study the peaks and pedestal in @xmath9 and @xmath8 and to discuss the physical origins of these jet characteristics .
our approach to the problem of correlation between particles in jets will be based on parton recombination that has been shown to be successful in reproducing the single - particle distributions in au+au collisions @xcite , and in d+au collisions @xcite .
it has also been applied to the study of dihadron correlations @xcite . however , since those investigations were carried out in the framework of the one - dimensional ( 1d ) formulation of the recombination model , a generalization to account for the 3d geometry of the problem is necessary if we are to describe the features in the @xmath1 and @xmath0 variables .
once we consider the 3d aspects of jets the angular distribution of the shower partons in the jet cone becomes an important new feature .
our treatment will be phenomenological , since no results from perturbative qcd considerations can reliably be applied to the intermediate @xmath4 region , which is where the data on correlations are taken .
furthermore , the pedestal phenomenon suggests that the jet interacts with the environment in a way that is sensitive to the hydrodynamical expansion of the medium , since there is longitudinal expansion , but not azimuthal expansion .
it turns out that we can relate the pedestal to the local enhancement of the thermal bath due to the energy loss of the hard parton to the medium .
our approach to the problem , though not based on first principles , provides the first theoretical interpretation of the pedestal and peak structure of the jets produced at rhic in quantitative terms .
the roles that shower and thermal partons play in the hadronization process are found to be crucial and can not be reinterpreted in any other model that we are aware of . in sec . 2 we describe how the 3d properties of jets are incorporated in the recombination formalism .
3 the pedestal phenomenon is investigated , followed by the study of the peaks in @xmath2 and @xmath3 distributions in sec . 4 . the conclusion is given in the last section .
in the framework of parton recombination formulated in 1d for particle production in hic @xcite-@xcite , let us start by writing down the single - and two - particle distributions @xmath10 @xmath11 where the recombination function ( rf ) for a pion is @xcite @xmath12 the two- and four - parton distributions , @xmath13 and @xmath14 , can be written in terms of their components as @xmath15 @xmath16 the thermal parton distribution is @xmath17 and the shower partons in a jet have the form @xmath18 where @xmath19 is the distribution of hard parton @xmath20 in a hic , and @xmath21 is the average fraction of hard partons that emerge from the thermal medium to hadronize .
@xmath22 is the shower - parton distribution ( spd ) of parton @xmath23 in a shower initiated by a hard parton @xmath20 ; their parameterizations for various @xmath20 and @xmath23 are given in @xcite .
for the sake of clarity we have omitted the labels for parton species for @xmath24 and @xmath25 in the above equations and in what follows , except when the identity of a parton is needed for emphasis .
it is important to recognize that eq .
( [ 7 ] ) should not be applied to every @xmath26 that appears in eqs .
( [ 4 ] ) and ( [ 5 ] ) because all the shower partons in @xmath13 and @xmath14 are created in the same jet , so all the @xmath27 terms share the same hard parton with momentum @xmath28 .
thus for two shower partons @xmath29 , their joint distribution is @xmath30 where the quantity in the curly brackets has the form in terms of momentum fractions @xmath31 and @xmath32 @xmath33 . \label{9}\end{aligned}\ ] ] this is done to guarantee momentum conservation @xmath34 and to symmetrize the order of emission of the two partons . not only should eq .
( [ 8 ] ) apply to two adjacent shower partons , like in @xmath35 and @xmath36 , which recombine to form the pions at @xmath37 and @xmath38 , respectively , but also to two shower partons like in @xmath39 , which do not recombine with each other , but separately with thermal partons .
it is the structure in eq .
( [ 8 ] ) that gives rise to correlation between two shower partons in a jet , and thereby endows the detected hadrons at high @xmath4 with correlation . however , that is sufficient primarily in the case of considering only the correlation among momentum magnitudes in 1d . to generalize our consideration to 3d ,
let us focus our attention first on the @xmath40 components in eq .
( [ 5 ] ) that give the most important contribution to the trigger and associated particles in central collision . to be specific
let us further designate the trigger particle to be a @xmath41 and the associated particle to be a @xmath42 .
the 3d expression for eq .
( [ 2 ] ) in that case is @xmath43 where @xmath44 is the 3d version of the rf and it includes the spatial coordinates .
since recombining partons must not only have collinear momenta but also have overlapping wave functions , there are many narrow gaussian distributions in @xmath44 that reduce the phase space of integration .
let us first restrict our attention to the @xmath45-@xmath46 plane , i.e.the plane containing the hard parton momentum @xmath47 and the longitudinal direction @xmath48 , and assume that all @xmath49 vectors are in that plane so that we may write @xmath50 and @xmath51 we further assume that @xmath52 is in the direction of @xmath47 and that the trigger momentum @xmath53 is also along @xmath47 . this is a simplification that does not compromise the angular correlation between the trigger and associated particles , which we shall study in detail .
since @xmath44 requires the wave functions of the recombining partons to overlap , the shower and thermal partons at @xmath54 and @xmath55 should be nearby .
these assumptions allow us to reduce the trigger momentum part of eq .
( [ 10 ] ) to the usual 1d formulation @xmath56 where @xmath57 is the usual rf in 1d , given in eq .
( [ 3 ] ) .
the superscript @xmath58 on @xmath24 is omitted on the rhs since the flavor dependence of the thermal partons is negligible and ignored .
we shall assume that at midrapidity , @xmath59 , the thermal distribution does not contain essential dependence on @xmath60 so that the condition @xmath61 does not lead to any restriction that would cause @xmath62 to deviate from the usual parameterization given in eq .
( [ 6 ] ) . in the case of the associated particle we must recognize that the shower parton @xmath63 may be emitted at an angle @xmath64 relative to the hard parton momentum @xmath47 . since @xmath47 forms an angle @xmath65 with the @xmath46 axis ( after identifying the direction @xmath66 with @xmath67 )
, we have @xmath68 the recombination of the shower parton at @xmath63 with a thermal parton at @xmath69 requires that they are not only collinear , but also overlapping spatially .
there are therefore @xmath70-functions , @xmath71 , that force the momentum @xmath72 of the associated particle to be in the same direction as @xmath73 and @xmath74 .
the corresponding 3d integrations result in @xmath75 this equation is very similar to eq .
( [ 13 ] ) but with one crucial difference ; that is , the shower parton distribution now depends on @xmath64 . in eq .
( [ 13 ] ) no angular dependence of the shower parton @xmath52 is assumed for simplicity , since it is only the difference between @xmath52 and @xmath63 that matters . that angular difference , @xmath64 , made explicit in eq .
( [ 14 ] ) , is translated into the angular difference between @xmath53 and @xmath72 .
we have tacitly assumed that the thermal partons have local angular spread that allows a @xmath76 to recombine with @xmath52 to form @xmath53 , and similarly @xmath69 with @xmath63 to form @xmath72 , in such a way that @xmath77 and @xmath78 are in the directions of @xmath79 and @xmath80 , respectively .
the dependence of a spd on @xmath64 can not be determined from pqcd , since the values of @xmath4 concerned is in the @xmath81 gev / c range .
we shall assume that the spd in eq .
( [ 15 ] ) can be written in the following factorizable form @xmath82 where the dependence on @xmath64 has a gaussian distribution @xmath83 \ , \label{17}\end{aligned}\ ] ] whose half - width depends on the momentum fraction @xmath45 of the shower parton @xmath84 with @xmath85 being a parameter that is adjustable .
@xmath86 on the rhs of eq .
( [ 16 ] ) is to be symmetrized with @xmath87 in eq.([13 ] ) without further complication from angular consideration .
the momentum fraction relevant for @xmath88 is @xmath89 , independent of @xmath90 .
once we obtain eq .
( [ 16 ] ) where @xmath91 stands as a modifying factor , it is reasonable to liberate that factor from the way in which it is derived , and allow it to assume a 3d property so that when we later consider the @xmath3 behavior in the transverse plane the same factor @xmath92 applies to the azimuthal angle between @xmath53 and @xmath72 with @xmath64 replaced by @xmath3 . using eqs .
( [ 12 ] ) and ( [ 14 ] ) we obtain with @xmath93 @xmath94 where @xmath95 is to be identified with @xmath96
. we now can write eq .
( [ 10 ] ) with the help of eqs .
( [ 3 ] ) , ( [ 13 ] ) and ( [ 15 ] ) in the simpler form @xmath97 where @xmath20 is summed over all hard parton species .
if @xmath20 is a valence quark of @xmath41 or @xmath42 , the corresponding spd is @xmath98 , while if @xmath20 is a sea quark , the corresponding spd is @xmath99 in the notation of @xcite , where @xmath100 and @xmath101 . the integral over @xmath28
will , in practice , be from 3 to 30 gev / c , and @xmath102 will be integrated from @xmath103 to @xmath104 , although the formalism is not reliable for any transverse momentum less than 1 gev / c .
there are other terms of @xmath14 contained in eq.([5 ] ) that can be written out as in eq .
( [ 20 ] ) , but will not be detailed here .
the experimental procedure of making background subtraction involves multiple considerations , and the result is that there remains a residual pedestal in @xmath8 in central collisions , but no pedestal in @xmath105 , where @xmath8 and @xmath9 are differences in @xmath0 and @xmath1 between the associated particle and the trigger . in @xcite
the dihadron correlation is calculated with the assumption that the factorizable part of the two - particle distribution corresponds to the background . that leaves only the @xmath106 terms of @xmath14 to consider .
we now find that in order to understand the origin of the pedestal what constitutes the background must be reconsidered .
the first point that we want to address is the effect of energy loss of hard partons traversing the hot medium .
the parameter @xmath21 in eqs.([7 ] ) and ( [ 8 ] ) , being only 0.07 as determined in @xcite for central au+au collisions , implies that a large fraction of the hard partons produced are absorbed or attenuated by the bulk medium .
those that emerge to hadronize outside must have been created near the surface . in the short distance
that such a parton travels in the medium it must on average lose some energy and locally enhance the thermal motion of the partons in the environment .
those enhanced thermal partons should have an effective inverse slope @xmath107 that is slightly higher ( by @xmath108 ) than that of the usual thermal partons not influenced by the passage of any hard partons .
the latter corresponds to what is measured for the low-@xmath4 pions in single - particle distribution , which for @xmath109 gev / c has led to the determination of @xmath110 gev / c in @xcite . if @xmath108 is small , as we expect it to be , it does not make too much difference in the calculation of the single - particle distribution in the intermediate @xmath4 region .
however , in principle , in thermal - shower recombination the shower partons recombine with the enhanced thermal partons that are in the immediate vicinity of the hard parton that creates the shower . to distinguish the two values of the inverse slope ,
let us use @xmath107 to denote the one for the enhanced thermal parton , since that is the only physically relevant one for @xmath111 recombination , i.e. , it is what appears in eq .
( [ 6 ] ) for @xmath112 .
let us use @xmath113 to denote the inverse slope for the thermal medium undisturbed by jet quenching , i.e. , @xmath114 gev / c .
we define @xmath115 which is a parameter to be determined phenomenologically in this paper , although it can presumably be determined directly by dedicated experimentation
. we shall relate @xmath108 to the pedestal .
the physics involved is sensible , since the pedestal is seen only for central collisions where @xmath111 recombination is important . in peripheral collisions
thermal partons play a minor role and there are few of them to be enhanced .
we now give an argument why the pedestal is seen in @xmath8 , but not in @xmath9 .
when a hard parton is first scattered or created at a large angle relative to the incident beam direction , the bulk medium is in a highly compressed state . as it expands longitudinally , the region of enhanced thermal partons expands with the whole system , although limited to the neighborhood of where the hard parton traverses the thin layer close to the surface .
thus a section at midrapidity can have the enhanced @xmath107 . in the transverse plane ,
on the other hand , the expansion is in the radial direction only ; that is , there is no mixing across different azimuthal sections .
the subtraction scheme carried out by star @xcite defines the yield @xmath116 to be zero at @xmath117 after subtraction .
since there is no expansion in the @xmath1 direction , we can take this subtraction scheme to imply that all thermal partons that are enhanced stay inside the @xmath118 region , and that for @xmath119 there is no thermal - thermal recombination that is not in the background .
there may still be an enhancement of the yield inside the @xmath118 region that plays a role similar to the pedestal in @xmath8 , but it can not extend beyond @xmath120 by virtue of the subtraction scheme .
the above description can not be demonstrated by a transport model , since we do not have a monte carlo code for the evolution process .
however , the physical content of our discussion will be embodied in our quantitative formulation below .
the recombination model actually provides a more transparent description of the hadronization process than what a code without reliable equations can offer .
let us first consider the particles that are associated with a trigger . since in au+au collisions
the @xmath111 component is dominant in the intermediate @xmath4 region @xcite , which includes the trigger window @xmath121 gev / c , we select the @xmath122 term in eq .
( [ 5 ] ) for our discussion below , although both terms in @xmath123 are included in our calculation .
thus the parts of @xmath14 that contribute to the trigger and its associated particles are @xmath124 among the three terms on the right side it is clear that @xmath125 are directly related to the jet and contribute to the peak in the associated particle distribution ( apd ) , while @xmath126 does not involve the shower partons but can nevertheless contribute to the apd outside the peak .
it is therefore our prime candidate for the pedestal .
the absence of an obvious pedestal in the apd in @xmath105 is a consequence of the background subtraction scheme in the analysis of the experimental data , where the @xmath9 distribution at @xmath117 is identified to be the background . in our description of the apd the background corresponds to @xmath127 where @xmath128 is the thermal parton distribution in the absence of any hard partons , i.e. , @xmath129 the meaning of @xmath113
has already been discussed in connection with eq .
( [ 21 ] ) .
we have tacitly assumed that the normalization factor @xmath130 is unchanged from that in eq .
( [ 6 ] ) .
the experimental subtraction scheme implies that @xmath131 at @xmath117 .
we now define the apd with background subtracted to be generated by @xmath132 in which @xmath133 vanishes at @xmath134 , but need not be zero at @xmath118 .
equation ( [ 26 ] ) is our basic input in the calculation of the apd .
let us now write our two - particle distribution in a form suitable for application of the experimental cuts . in eq .
( [ 20 ] ) we have a differential distribution in @xmath135 where @xmath95 and @xmath136 denote the variables of the trigger , and @xmath0 and @xmath1 those of the associated particle . with the definition @xmath137
we change the differentiated variables to @xmath138 , with @xmath136 being used as a free reference point in the assemblage of the @xmath3 distribution ; the trigger @xmath95 is to be integrated over the trigger window @xmath139 .
we then have @xmath140 the experimental apd involves integrations over the trigger window in @xmath95 and @xmath37 and the window of the associated particle in @xmath38 .
it is @xmath141 where @xmath142 is the one - particle trigger distribution that receives contribution from the @xmath122 component only , i.e. @xmath143 the two - particle distribution @xmath144 in eq .
( [ 29 ] ) can be divided into two explicit pieces . for the remainder of this section
we consider only the piece related to the pedestal @xmath145\quad , \label{31}\end{aligned}\ ] ] where we have extracted the @xmath1 dependence of the associated particle in the form of a factor @xmath146 , which satisfies some constraints to be specified below .
it should be noted that eq .
( [ 31 ] ) is factorizable , since the two parts of @xmath147 are independent of each other .
the @xmath148 part contributes to the trigger distribution given in eq.([30 ] ) .
it will be cancelled in the ratio defined in eq .
( [ 29 ] ) . using eq .
( [ 24 ] ) , the integration of the last term in eq .
( [ 31 ] ) over @xmath149 can readily be carried out , giving @xmath150 which , upon further integration over the window @xmath151 of the associated particle , results in @xmath152 where @xmath153 similar result follows for the @xmath154 term in eq .
( [ 31 ] ) .
next , we consider the @xmath8 and @xmath9 dependences of the pedestal in eq .
( [ 31 ] ) . as we have discussed qualitatively in the beginning of this section
, the longitudinal expansion allows the enhanced thermal partons to extend over a wide range in @xmath0 , but radial expansion does not increase the range in @xmath1 . putting these properties into quantitative terms ,
we give @xmath155 no essential dependence on @xmath8 in eq .
( [ 31 ] ) , but a gaussian dependence on @xmath9 so that @xmath156 the normalization @xmath157 is to be determined by the condition that when @xmath3 is integrated over the experimental window @xmath158 we obtain @xmath159 in this way we can relate eq .
( [ 31 ] ) to the pedestal observed in the experiment without explicit factors dependent on the experimental windows .
the half - width @xmath160 is adjusted so that the quantity inside the square brackets in eq .
( [ 31 ] ) vanishes at @xmath134 , as required by the subtraction scheme , eq .
( [ 25 ] ) .
what we do here is to fix all the extra free parameters of the problem by the experimental features of the data . since our aim is to reproduce the observed characteristics of the data , which are presented with specific cuts , it is impossible to do so without incorporating those cuts .
however , it does not imply that we are merely fitting the data with free parameters .
we shall perform several multi - dimensional integrals to obtain in the next section the apd with the peak whose magnitude is a prediction of our model calculation . upon integration of eq .
( [ 29 ] ) over @xmath3 in the acceptance window , we get for the pedestal part @xmath161 \ , \label{37}\end{aligned}\ ] ] where we have made use of the factorizability of eq .
( [ 31 ] ) and the result of integration over @xmath38 given in eq .
( [ 33 ] ) .
this is the constant pedestal in @xmath8 , which we relate to @xmath108 . using @xmath162 ( gev / c)@xmath163 , @xmath114 gev / c from @xcite , @xmath164 gev / c as in the experiment @xcite , and eq.([21 ] ) for the definition of @xmath108
, we fit the pedestal height of @xmath165 @xcite and obtain @xmath166 thus the inverse slope of the enhanced thermal partons is only 5% higher than that of the un - enhanced partons , small enough to have a negligible effect on the calculation of the single - particle distribution using @xmath113 .
however , the effect on the associated particle distribution is evidently not negligible .
although eq .
( [ 38 ] ) is a fitted result , it should be emphasized that it is in the specific model of parton recombination that the pedestal phenomenon is interpreted as a consequence of the enhancement of the thermal energy of the soft partons due to energy loss of the hard parton
. equation ( [ 37 ] ) is the projection of @xmath167 onto @xmath8 .
we now project it onto @xmath9 by integrating it over @xmath8 from @xmath168 to @xmath169 in accordance to the experimental cut @xcite and obtain from eq .
( [ 31 ] ) for the pedestal part @xmath170 \quad . \label{39}\end{aligned}\ ] ] this is not a flat pedestal as in the case of @xmath8 , but is nevertheless a remnant of the effect of enhanced thermal partons sitting under a peak from the @xmath171 part of @xmath172 that we have not yet calculated .
although it is not visible as a plateau , its origin is the same as that which gives rise to the pedestal in @xmath173 .
it is the background subtraction in @xmath9 that forces eq .
( [ 39 ] ) to vanish at @xmath174 , which is achieved by our choice of @xmath160 in eq .
( [ 35 ] ) its value is @xmath175 with this choice of @xmath160 eq .
( [ 39 ] ) is entirely specified numerically .
the @xmath3 dependence of @xmath176 is a broad mount between @xmath177 , which we shall exhibit later together with the complete @xmath178 .
this mount is the projection of a ridge in the 3d display of apd in @xmath179 onto the @xmath3 subspace .
this completes our discussion of the @xmath180 contribution to @xmath172 in eq .
( [ 26 ] ) and then to eqs.([28 ] ) and ( [ 29 ] ) .
we now proceed to the last two terms of @xmath172 .
we now return to eq .
( [ 26 ] ) and consider the @xmath181 terms in @xmath172 that we have put aside .
the @xmath111 component has already been described in detail in eq .
( [ 20 ] ) .
the @xmath182 component is less important except in peripheral collisions .
it can be included by the replacement of the @xmath183 term in eq .
( [ 20 ] ) by a @xmath184 term with appropriate symmetrization .
however , since the recombination of two shower partons reproduces the fragmentation function in accordance to @xmath185 from which the spds are derived in the first place , we may replace a @xmath186 term by a @xmath187 term . for notational simplicity
we leave the @xmath36 terms out in our description below , but numerically include their recombination in the calculated result .
let us recall that starting from the general formula eq .
( [ 10 ] ) we first focussed on the variables in the plane containing @xmath47 and the longitudinal direction @xmath48 and related the relevant angles to pseudorapidities in eq .
( [ 12 ] ) .
we then defined the angle @xmath64 between the shower parton @xmath63 and the hard parton @xmath47 to be as given in eq .
( [ 14 ] ) .
the angular dependence of the spd is then expressed by a gaussian distribution @xmath188 in eq.([17 ] ) , put in a factorized form in eq .
. we may identify @xmath188 as the @xmath45-@xmath46 projection of a general gaussian distribution that describes the dependence on the angle between @xmath53 and @xmath72 in 3d geometry . in the small width approximation
, it can be shown that the corresponding projection onto the transverse , @xmath45-@xmath189 , plane , gives a gaussian form , @xmath190 , for the @xmath3 dependence .
returning to eq .
( [ 20 ] ) which is written for the plane containing @xmath47 and @xmath48 , we now write the double differential in both @xmath8 and @xmath9 , as for the pedestal term , but now for the peak term arising from the @xmath39 component @xmath191 where @xmath192 is a numerical normalization factor such that when eq.([42 ] ) is integrated over @xmath9 from @xmath193 to @xmath194 ( the experimental window for projection to the @xmath8 dependence ) one gets @xmath195 in the small @xmath45 approximation . identifying eq .
( [ 42 ] ) as @xmath196 , we can substitute it into the numerator of eq .
( [ 29 ] ) and obtain @xmath197 .
there is only one free parameters to adjust ; it is @xmath85 in eq .
( [ 17 ] ) .
the projections of the double differential distribution to @xmath8 and @xmath9 separately can be compared with the data when the pedestal component is added .
that is @xmath198 \quad , \label{44}\end{aligned}\ ] ] @xmath199 \quad .
\label{45}\end{aligned}\ ] ] the pedestal contributions to the above integrals are given by eqs .
( [ 37 ] ) and ( [ 39 ] ) , respectively .
it should be noted that the latter has a @xmath3 dependence described by @xmath200 , which is a gaussian with a half - width @xmath160 given by eq.([40 ] ) .
we shall find below that the half - width of the peak term , controlled by @xmath85 , is much smaller than @xmath160 , so the last terms in eqs .
( [ 44 ] ) and ( [ 45 ] ) dominate the peak structure of the apd .
the term @xmath201 gives rise to a ridge , whose projections are the flat pedestal in @xmath202 , and a broad mount in @xmath203 .
after we put all the pieces together , using eq .
( [ 29 ] ) as the important link between our calculation and the measurable quantities , the results from eqs .
( [ 44 ] ) and ( [ 45 ] ) can be compared to data with @xmath204 adjusted to fit the peak width .
figure 1 shows the apd in @xmath8 . the solid line is the result of our calculation when we set @xmath205 evidently , it reproduces the data @xcite satisfactorily , both in the peak structure and the pedestal .
it should be emphasized that although the width and the pedestal height are adjusted to fit by our choice of @xmath206 and @xmath108 , the height of the peak is a consequence of our very complicated calculation involving multiple integrals over many terms based on the recombination model .
it is by no means trivial that the data can be fitted so well . in point of fact
, we have only calculated @xmath207 production , whereas the data are for all charged particles .
thus the precise values of the parameters are not as significant as the overall situation where the pedestal and peak structure of the @xmath173 distribution of the data can be reproduced by our description of dihadron correlation .
turning now to the apd in @xmath9 , we have no free parameters to adjust , since the double differentials in the square brackets in eqs.([44 ] ) and ( [ 45 ] ) are identical . upon integration over @xmath173
we obtain the solid line in fig . 2 , which compares well with the data @xcite .
the dashed line indicates the pedestal contribution from eq .
( [ 39 ] ) .
it plays the role of the flat pedestal in fig . 1 ,
but here it vanishes at @xmath117 because of the subtraction scheme .
since the peak structure in @xmath9 is much narrower than the broad mount of the pedestal ( due to @xmath208 ) , we have applied the vanishing condition at @xmath117 only to the @xmath209 component , knowing that the @xmath210 component is negligible at the wings of the peak . figures 1 and 2 represent our main results on dihadron correlation in central au+au collisions at @xmath211 gev , when the trigger particle is kept within the range @xmath212 gev / c and the associated particle in the range @xmath213 gev / c .
we can readily calculate the apd at higher momentum ranges , but not at lower momenta , since our model is not reliable for parton momentum less than 1 gev / c .
for that reason we do not calculate the apd for @xmath214 gev / c , even though data for that range are presented in @xcite .
to summarize , we have successfully reproduced the data that show the existence of peaks in @xmath2 and @xmath3 distributions , the former sitting above a flat pedestal , while the latter sitting above a broad mount .
some parameters are used to fit the data , but the essence of our work is not data fitting .
we have demonstrated that the physics underlying the detail structure of the jet characteristics observed in rhic experiments can be understood in the framework of parton recombination .
jets produced in heavy - ion collisions create shower partons that are in the environment of thermal partons , which can themselves be enhanced by the passage of hard partons through the medium .
that view is probably shared by all theoretical approaches to the problem .
the issue then is how those partons hadronize .
different models treat the hadronization process differently .
our approach emphasizes the recombination of the thermal partons and the shower partons in the intermediate @xmath4 region . that is our way of accounting for the medium effects on jets .
so far we have not encountered any obstacle in understanding the data in that way .
data fitting is only a concrete demonstration that the details of the jet structure can be quantitatively reproduced in our treatment .
the hadronization formalism is thereby enriched by the determination of some features in the model by phenomenology .
our first discovery in this work is that the pedestal in the @xmath2 distribution can be related to @xmath108 in the local thermal distribution .
no long - range correlation " has been put in by hand .
the pedestal is a consequence of the recombination of the thermal partons among themselves , which are only indirectly affected by hard scattering through the enhancement of @xmath107 due to energy loss .
thus the pedestal is not a part of the jet , but can not be present without a jet .
this chain of successive connections involving energy loss , enhanced thermal partons , their hadronization by recombination , and the elevated apd in @xmath2 is certainly very different from a model where an explicit correlation is put in by hand .
our second achievement in this work is the success in describing the peak structures in both @xmath2 and @xmath3 distributions in terms of one angular distribution of the shower partons in a jet cone .
short - range correlation " has been put in by hand .
the only primitive correlation in the problem is that which exists among the shower partons in a jet .
the properties of that correlation in the momentum magnitudes of the shower partons have already been described in @xcite . here
we show how the correlations manifest themselves in the angular variables of the produced pions .
there is one caveat in our analysis that should be noted .
we have calculated only @xmath207 production , whereas the data are on all charged hadrons .
thus numerically the parameters determined here are not definitive .
since particle identification is steadily being improved , the more appropriate arena for detailed matching of theory and experiment is when the data for the production of specific species become available . despite the tentativeness of the value of @xmath108 given in eq.([38 ] ) on account of the statement made just above , our assertion that the pedestal is related to the enhancement of @xmath107 of the thermal partons remains unaffected . as a test of that assertion
we propose that the proton associated with a @xmath41 trigger be measured in addition to the measurement of a @xmath41 or @xmath42 associated particle .
the @xmath215 ratio of the associated particles in a @xmath41 triggered jet should be a good probe of the physical mechanism underlying both the pedestal and the peak .
the recombination approach to the problem is well positioned to calculate that ratio , as it has for single - particle distribution @xcite .
we surmise that the pedestal part of the @xmath215 ratio would be higher than in the single - particle case because @xmath216 .
what we have done in this paper is influenced greatly by our intention to understand the pedestal and peak structure in the data of [ 3 ] . as a consequence ,
our analysis involves many integrals that correspond to the experimental cuts in the data .
having determined the origin of the pedestal and peaks in the @xmath2 and @xmath3 distributions , we are now equipped to launch a study of the correlation problem independent of any triggers and related experimental cuts , on which more and more data are becoming available .
we have benefitted from discussions with x. n. wang and c. b.yang in the preliminary stage of this work .
communication with f.wang has also been helpful .
this work was supported , in part , by the u. s. department of energy under grant no .
de - fg03 - 96er40972 .
99 for reviews see m. gyulassy and l. mclerran , nucl .
a * 750 * , 30 ( 2005 ) .
p. jacobs and x. n. wang , prog . part .
phys . * 54 * , 443 ( 2005 ) .
j. adams et .
al . , ( star collaboration ) , nucl - ex/0501016 .
c. a. gagliardi ( for star collaboration ) , proceedings of hard probes 04 , ericeira , portugal ( 2004 ) .
d. magestro ( for star collaboration ) , proceedings of hard probes 04 , ericeira , portugal ( 2004 ) .
y. guo ( for star collaboration ) , nucl - ex/0502015 .
b. jacak ( for phenix collaboration ) , proceedings of hard probes 04 , ericeira , portugal ( 2004 ) .
j. adams _ et al . _
( star collaboration ) , nucl - ex/0411003 .
r. c. hwa and c. b. yang , phys .
c * 70 * , 024905 ( 2004 ) .
r. c. hwa and c. b. yang , phys .
lett . * 93 * , 082302 ( 04 ) .
r. c. hwa and c. b. yang , phys .
c * 70 * , 054902 ( 2004 ) .
r. c. hwa and z. tan , nucl - th/0503052 .
r. c. hwa and c. b. yang , phys .
c * 66 * , 025205 ( 2002 ) .
r. c. hwa and c. b. yang , phys .
c * 70 * , 024904 ( 2004 ) .
fig . 1 . associated particle distribution in @xmath2 for @xmath217 gev / c with trigger particle in @xmath218 gev / c . the data from ref .
@xcite are for all charged hadrons in the respective @xmath4 ranges .
the solid line is the result of our calculation of @xmath42 associated with @xmath41 trigger .
fig . 2 .
same as in fig . 1 except that the distribution is in @xmath3 .
the dashed line represents the pedestal effect in @xmath3 forced to vanish at @xmath219 by the subtraction scheme discussed in sec . 3 . | we study the characteristics of correlation between particles in jets produced in heavy - ion collisions . in the framework of parton recombination
we calculate the @xmath0 and @xmath1 distributions of a pion associated with a trigger particle .
the origin of the pedestal in @xmath2 is related to the longitudinal expansion of the thermal partons that are enhanced by the energy loss of hard partons traversing the bulk medium .
the peaks in @xmath2 and @xmath3 are related to the same angular spread of the shower partons in a jet cone .
no artificial short- or long - range correlations are put in by hand .
a large part of the correlation between hadrons in jets is due to the correlation among the shower partons arising from momentum conservation .
recombination between thermal and shower partons dominates the correlation characterisitics in the intermediate @xmath4 region . 0.5 cm pacs numbers : 25.75.gz , 25.75.-q = 6.5 in = -0.5 in .75 cm * charles b. chiu@xmath5 and rudolph c. hwa@xmath6 * .5 cm @xmath5center for particle physics and department of physics + university of texas at austin , austin , tx 78712 , usa + @xmath7institute of theoretical science and department of physics + university of oregon , eugene , or 97403 - 5203 , usa |
millimeter and submillimeter observations of dense cores in nearby parsec - scale cluster - forming clumps have shown that the core mass function ( cmf ) resembles the stellar initial mass function ( imf ) .
for example , @xcite performed 1.2 mm continuum observations toward l1688 and identified 57 starless dense cores whose mass function ( @xmath0 ) is consistent with the salpeter power - law imf at the high mass end ( @xmath1 ) .
the cmf also has a break at about 0.3 m@xmath2 , below which it flattens to @xmath3 .
other authors have found similar cmfs in the region ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the shape of the cmf is broadly consistent with the imf of class ii ysos in the same region @xcite .
another example is the serpens cloud core , where @xcite identified , through 3 mm continuum interferometric observations , 32 dense cores whose mass function can be fitted by a power - law of @xmath4 , again consistent with the salpeter imf . these and other observations suggest that the identified dense cores may be the direct progenitors of individual stars and the bulk of the stellar imf may be at least partly determined by cloud fragmentation in the parsec - scale dense clumps .
thus , understanding the formation process of dense cores is a key step towards a full understanding of how stars form .
both supersonic turbulence and magnetic fields are expected to play a role in clustered star formation .
the relative importance of the two is still under debated , however .
one school of thought is that the core ( and thus star ) formation in a cluster - forming clump is mainly regulated by supersonic turbulence . in this scenario ,
star formation is completed rapidly , before the initial supersonic turbulence has decayed significantly ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
magnetic fields are less important in this picture , although @xcite argued that a weak magnetic field is still needed in order to produce a core mass spectrum that resembles the stellar imf .
the weak magnetic field is expected to be strongly tangled by the supersonic turbulence . ordered field structures , if present at all , are typically attributed to large - scale compression and are expected to be parallel to the dense filaments that are also produced by the same compression ( e.g. , * ? ? ?
the second school of thought envisions a magnetic field that is more dynamically important ; it prevents stars from forming too rapidly . because supersonic turbulence dissipates rapidly ,
it must be somehow replenished . in a cluster - forming clump
, protostellar outflow feedback can play an important role in turbulence regeneration ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
@xcite demonstrated that protostellar outflows can resupply the supersonic turbulence , keeping the clumps near a quasi - virial equilibrium state for a relatively long time ( see also * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
because the matter moves preferentially along the dynamically important magnetic field , the field is expected to be more or less perpendicular to the dense filaments that are created by either turbulence compression or self - gravity .
a goal of this paper is to quantify the effects of the magnetic field on the clustered star formation in general , and the properties of dense cores in particular , through 3d mhd simulations that include outflow feedback for turbulence replenishment .
the rest of the paper is organized as follows . in section [ sec :
model ] , we describe the numerical method and simulation setup . in section [ sec : results ] , we apply a version of the so - called `` clumpfind '' algorithm to identify dense cores in a parent dense clump and derive their properties .
we find that the internal motions of the cores are sensitive to the magnetic field strength of the parent clump and that the external surface pressures play a dominant role in core formation in outflow - driven turbulence . in section [ sec : discussion ]
, we compare the properties of the simulated cores with those observed in a couple of nearby cluster forming regions .
our main conclusions are summarized in section [ sec : summary ] .
the initial and boundary conditions of the simulations are essentially the same as those of @xcite .
we consider a centrally - condensed molecular gas clump inside a cubic simulation box of @xmath5 pc on each side , with an initial density profile of @xmath6^{-1}$ ] for @xmath7 ( where @xmath8 is the radius of the central plateau region ) and @xmath9 for @xmath10 . here , the central density @xmath11 is given by @xmath12 g @xmath13 , with @xmath14 being the central number density of molecular hydrogen , assuming 1 he for every 10 h atoms .
periodic boundary condition is applied to each side of the box .
we adopt a central h@xmath15 density of @xmath16 @xmath13 , corresponding to a central free - fall time @xmath17 myr and a central jeans length of @xmath18 pc .
it yields a total clump mass of @xmath19 .
the average clump density is @xmath20 @xmath13 , corresponding to a global free - fall time @xmath21 myr .
an isothermal equation of state is assumed , with a sound speed of @xmath22 km s@xmath23 for gas temperature @xmath24 20 k. our choice of the clump mass , density and gas temperature is motivated by observations of the nearest pc - scale cluster forming clump , the @xmath25 ophiuchi main cloud . from @xmath26co ( @xmath27 ) observations , @xcite estimated a total gas mass of 865 @xmath28 for the whole l1688 region , adjusted for the parallax distance of 125 pc ( e.g. , * ? ? ?
* ; * ? ? ?
the gas temperature in the region has some spatial variation , but the average appears close to 20 k. our clump parameters are also consistent with those of other nearby cluster - forming clumps @xcite and the infrared dark clouds which are thought to be future sites of cluster formation in gmcs ( e.g. , * ? ? ?
at the beginning of the simulation , we impose on the clump a uniform magnetic field along the @xmath29-axis .
the field strength is specified by the plasma @xmath30 , the ratio of thermal to magnetic pressures at the clump center , through @xmath31 @xmath32 g . in units of the critical value @xmath33 @xcite ,
the mass - to - flux ratio in the central flux tube is given by @xmath34 .
the mass - to - flux ratio for the clump as a whole is @xmath35 .
although systematic measurements of magnetic field strength in cluster - forming clumps are not available , relatively strong magnetic fields are inferred in some cases : e.g. , 850 @xmath32 g for omc 1 at @xmath36 @xmath13 ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and 160 @xmath32 g for serpens at @xmath37 @xmath13 @xcite . in the present paper , we adopt as a representative value @xmath38 ( or 100 @xmath32 g at @xmath39 @xmath13 ) , corresponding to @xmath40 or a dimensionless flux - to - mass ratio @xmath41 , so that the clump as a whole , as well as the denser central region , is magnetically supercritical . following the standard practice ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , we stir the initial clump at the beginning of the simulation with a turbulent velocity field of power spectrum @xmath42 and rms mach number @xmath43 .
our initial clump has a virial parameter ( the ratio of the kinetic energy to the gravitational energy ) of @xmath44 .
the effective @xmath45 should be close to unity for @xmath46 since the gravitational energy is effectively reduced by the magnetic field by a factor of @xmath47 .
the turbulence is allowed to decay freely , except for feedback from protostellar outflows .
the evolution of the turbulent , magnetized molecular clump is followed using a three - dimensional mhd code based on an upwind tvd scheme .
the mhd code is essentially the same as those used in @xcite and @xcite .
the ideal mhd equations are solved using the code having second - order accuracy in both space and time .
the equation of state is assumed to be isothermal . to ensure the divergence - free magnetic field ,
the divergence cleaning method is adopted . the poisson equation for gravitational potential
is solved using the fast fourier transform .
our simulation has a resolution of @xmath48 .
although our focus is on core properties , we do include a crude treatment of the formed stars and outflows , following @xcite ; we refer the reader to that paper for details .
briefly , when the density in a cell crosses the threshold @xmath49 , corresponding to @xmath50 @xmath13 , we create a lagrangian particle at the cell center .
we extract mass from a small region surrounding the cell , and put it on the particle , which moves with mass - weighted mean velocity of the extracted mass ( see nakamura & li 2007 for more detail ) .
after creation , the particle is allowed to accrete from the surrounding gas according to the same prescription as @xcite . to mimic the effect of protostellar outflows , the particle injects into the ambient gas a momentum that is proportional to the particle mass increment @xmath51 .
the outflow momentum is scaled with the dimensionless outflow parameter @xmath52 as @xmath53 , where @xmath54 is the wind velocity ( see the appendix of * ? ? ?
the fiducial value of @xmath52 is set to 0.4 , consistent with @xcite .
each outflow has a bipolar and spherical component , with a momentum ratio of 3:1 .
the model parameters are summarized in table [ tab : model ] .
we first concentrate on the dense cores identified in three representative models : model n1 ( no magnetic field , @xmath55 ) , w1 ( weak magnetic field , @xmath56 ) , and s1 ( moderately - strong magnetic field , @xmath38 ) .
their dimensionless outflow parameter is set to the fiducial value @xmath57 .
following @xcite , we identify dense cores from the density data cube using a variant of the clumpfind algorithm of @xcite .
we adopt a threshold density @xmath58 ( @xmath59 @xmath13 ) , and a maximum density @xmath60 ( @xmath61 @xmath13 , above which a lagrangian particle is created ) .
thus , the cores identified from our simulation data correspond roughly to the dense cores observed in the dust continuum emission and high density molecular tracers such as n@xmath15h@xmath62 ( @xmath27 ) and h@xmath26co@xmath62 ( @xmath27 ) .
the density distribution between @xmath63 and @xmath64 is divided into 10 bins equally spaced logarithmically .
we include only those spatially resolved cores containing more than 50 cells in the analysis , to ensure that their properties are determined with reasonable accuracy .
we also choose the cores that do not include particles .
in this sense , our identified cores correspond to starless cores .
the minimum core mass and radius identified through this procedure are @xmath65 and @xmath66 pc , respectively .
we note that our core identification procedure is not exactly the same as those used in observational studies , which are typically based on position - position - velocity data cubes of molecular line emission or column density maps from dust continuum emission . before discussing the physical properties of the identified cores ,
we show how the magnetic field affects the global density distribution .
figure [ fig : snapshot ] compares the column density distributions along the @xmath67-axis for the three models with different initial magnetic field strengths , at a stage when the star formation efficiency has reached @xmath68 .
since the star formation is retarded by the magnetic field , the evolution time tends to be longer at a given sfe for a stronger magnetic field . for comparison ,
the positions of the formed stars and the identified dense cores projected on the @xmath69 plane are indicated by the dots and triangles , respectively , in fig .
[ fig : snapshot ] .
the initial magnetic field direction is parallel to the @xmath29-axis ( the abscissa ) .
figure [ fig : snapshot ] indicates that the global density distribution depends on the initial magnetic field strength . for the non - magnetic model ( model n1 )
, the global density distribution shows a large - scale filamentary structure that contains many small fragments ( figs .
[ fig : snapshot]a and [ fig : snapshot]d ) .
the formation of the large - scale filament is probably because the turbulence energy is initially highest on the largest scale and the large - scale compression may be further amplified by self - gravity to create the filamentary structure . for the weakly - magnetized case ( model w1 )
, the global density distribution does not clearly show a large - scale filamentary structure by the stage shown in figs .
[ fig : snapshot]b and [ fig : snapshot]e .
the dense part appears to contain many smaller filaments or elongated fragments that are distributed almost independently of the initial magnetic field direction .
this is different from the non - magnetic case .
the difference may be because the local turbulent motions amplify random magnetic field components ( see section [ subsec : amplification ] ) , which slow down the formation of the large - scale filamentary structure .
in contrast , in the presence of a moderately - strong magnetic field ( model s1 ) , the cloud material condenses preferentially along the magnetic field lines into a large - scale filamentary structure that is nearly perpendicular to the initial magnetic field direction .
several less - dense filaments are also associated with the large - scale filament ( hereafter the main filament ) .
the less - dense filaments appear to be more or less along the initial magnetic field direction and to merge toward the dense parts of the main filament .
the dense cores are distributed primarily along the main filament .
the column density distribution is reminiscent of the hub - filament structures of star forming regions pointed out by @xcite .
the observed hub - filament structures may indicate the presence of the moderately - strong , spatially - ordered magnetic field , instead of the weak magnetic field that are tangled by supersonic turbulence .
we will revisit this issue in section [ subsec : polarization ] , where we compare dust polarization maps derived from the simulation data for models w1 and s1 . as shown in @xcite and @xcite ,
our model clump has reached a quasi - equilibrium state within one global free - fall time .
after that , the statistical properties of the cores remain largely - unaltered . therefore , to compare the statistical properties of the identified cores , we choose the same stage as that of fig .
[ fig : snapshot ] , i.e. , the stage when 16 % of the total mass has been converted into stars . in this subsection
, we compare the core properties for the three representative models , models n1 , w1 , and s1 .
the total numbers of identified cores are 160 , 105 , and 105 cores for models n1 , w1 , and s1 , respectively .
we now focus on the physical properties of the cores identified in the three representative models . to facilitate comparison with observations , we present the core properties in dimensional units .
figures [ fig : hist]a through [ fig : hist]i show the histograms of the core radius , mass , and velocity dispersion of the nonthermal component . the minimum , maximum , mean , and median values
are summarized in table [ tab : physprop ] .
the core radius , @xmath70 , is defined as the radius of a sphere having the same volume as the core .
the distributions of core radius are broadly similar in all cases .
they range from @xmath710.01 pc to @xmath720.05 pc , all peaking around 0.02 pc . the maximum and mean core radii are somewhat larger in model s1 ( the moderately - strong field case ) than in the other two .
the core mass distribution shows a similar trend . in all models ,
the core mass ranges from @xmath71 0.07 @xmath28 to @xmath73 .
the mean value is somewhat larger for model s1 than for the other two .
the mean core mass and radius are larger in model s1 probably because the stronger magnetic field can support more mass against gravitational collapse .
the magnetic effect is even more prominent in the distribution of core velocity dispersion . for the non - magnetic model ( model n1 ) , the velocity dispersion ranges from @xmath71 0.1 km s@xmath23 to 4 km s@xmath23 , peaking around 1 km s@xmath23 ( corresponding to a mach number of @xmath72 4 ) .
most of the cores have supersonic internal motions in this case , which appears inconsistent with the predominantly subsonic or at most transonic internal motions inferred observationally ( e.g. , andre et al . 2007 ; walsh et al . 2007 ; maruta et al .
2010 , see also section [ subsec : observations ] ) . even a relatively weak magnetic field appears to affect the internal motions inside the cores significantly . for model w1
, the median velocity dispersion is estimated at @xmath710.5 km s@xmath23 , corresponding to a mach number of about 2 , although the mean is almost the same as that of model n1 .
most of the cores in this case still have supersonic velocity dispersions .
in contrast , for model s1 , the majority of the cores have a velocity dispersion close to , or less than , @xmath71 0.3 km s@xmath23 , which is comparable to the sound speed ( @xmath74 km s@xmath23 ) .
the velocity dispersion ranges from @xmath71 0.05 km s@xmath23 to 2 km s@xmath23 , peaking around the median of 0.3 km s@xmath23 .
in other words , the strong magnetic field tends to reduce the core internal motions significantly .
the reduction is probably because the magnetic field cushions the converging flows in the supersonic turbulence that are largely responsible for the core formation .
the magnetic effect is also evident in the virial state of the cores , as we show below .
the internal motions of the observed dense cores are often measured using the velocity dispersions derived from molecular line data . here
, we compare the velocity dispersions for the three representative models with different field strengths to study the effects of the magnetic field .
figure [ fig : velocity - radius ] plots the non - thermal components of 3d velocity dispersions of the identified cores against the core radii for ( a ) model n1 , ( b ) model w1 , and ( c ) model s1 .
the best - fit power - laws are given by @xmath75 ( @xmath76 ) , @xmath77 ( @xmath78 ) , and @xmath79 ( @xmath80 ) for models n1 , w1 , and s1 , respectively , where @xmath81 denotes the correlation coefficient .
it shows that there is no clear correlation between the velocity dispersion and core radius .
our cores do not follow the power - law linewidth - radius relation derived by @xcite .
the same conclusion holds for dense cores formed in even more strongly magnetized , magnetically subcritical , clouds ( see figure 18 of * ? ? ?
the reason is probably that the dense cores created out of turbulent molecular gas are generally not in virial equilibrium ( the equilibrium is implied by larson s laws ) and that their formation is strongly influenced by external turbulent flows ( see below ) .
quantitatively , the velocity dispersions of the cores depend on the initial magnetic field strength , as shown already in figs .
[ fig : hist]g through [ fig : hist]i . in the absence of magnetic field ( model n1 ) , about 40% of the cores have velocity dispersions larger than 1 km s@xmath23 ; the majority of the cores have supersonic internal motions .
as mentioned above , even a relatively weak magnetic field can influence the internal motions significantly . in model w1 , although many cores still have large velocity dispersions , the median value is less than that in model n1 by a factor of @xmath82 . in the stronger field case ( model s1 ) , the median and mean of the velocity dispersions are estimated at 0.31 and 0.46 km s@xmath23 .
about 40 % of the cores have velocity dispersions smaller than the sound speed of 0.266 km s@xmath23 ; the majority of the cores have subsonic or at most transonic internal motions , which is more consistent with observations .
we conclude that the magnetic field plays a significant role in reducing the internal motions of the cores .
the virial parameter , the ratio of the virial mass to core mass , is often used as a measure of the gravitational boundedness of dense molecular cloud cores , particularly in observational studies . in fig .
[ fig : virial parameter ] , we plot the virial parameters against the core mass for the three models . here , we calculate the virial parameter as @xmath83 where @xmath84 is the 1d fwhm velocity width including thermal contribution and the virial mass of each core is obtained by assuming that it is a uniform sphere .
the virial mass depends on the density distribution . for a centrally condense sphere with @xmath85 , the virial parameter
is reduced by a factor of @xmath86 .
note that the effects of the nonspherical mass distribution appear to be small as long as the aspect ratios of the cores are not far from unity @xcite .
figure [ fig : virial parameter ] indicates that the relationship between the virial parameter and core mass depends on the initial magnetic field strength . in the absence of a magnetic field ( model n1 ) , the virial parameters are typically very large , and have a large scatter .
the scatter in the virial parameter - mass plot comes mainly from the large scatter in the velocity dispersion - radius relation ( see fig . [
fig : velocity - radius ] ) .
the virial parameters often exceed @xmath87 even for relatively massive cores , indicating that , for most of the cores , the gravitational energy is much smaller than the internal kinetic energy .
the virial parameters tend to increase with decreasing core mass , and the lower bound of the virial parameters follows roughly a power - law relation of @xmath88 . for the model with a weak magnetic field ( model w1 ) , the virial parameter - mass relation is qualitatively similar to that of the non - magnetized model ( model n1 ) .
although the fraction of cores with large virial parameters is somewhat small for the weak magnetic field case , the scatter in the viral parameter - mass relation remains large .
about 30 % of the cores still have virial parameters larger than 10 . on the other hand , for the model with a moderately - strong magnetic field ( model s1 ) , the identified cores tend to follow a single power - law relation of @xmath88 , with a rather small scatter . furthermore , most of the cores ( about 9 % of the cores ) have virial parameters smaller than 10 . this trend is consistent with the initially - magnetically subcritical case ( see figure 17 of * ? ? ?
it is also in good agreement with the scaling found by @xcite for nearby giant molecular clouds .
according to @xcite , an object confined by the ambient pressure has a virial ratio that is proportional to @xmath89 , where @xmath90 is the jeans mass defined by equation ( 2.13 ) of @xcite .
the fact that our cores follow the same relation implies that the ambient pressure plays an important role in confining our cores as well . in the next subsection
, we perform a detailed virial analysis of the cores to quantitatively clarify the role of the surface pressure term in core dynamics .
the virial theorem is useful for analyzing the core dynamical properties . here
, we perform a detailed virial analysis using the virial equation in eulerian coordinates ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) @xmath91 where @xmath92 , @xmath93 , @xmath94 , @xmath95 , @xmath96 , @xmath97 \cdot \mbox{\boldmath $ ds$}$ ] , and @xmath98 .
the quantities @xmath99 , @xmath100 , @xmath101 , @xmath102 , @xmath103 , @xmath104 , @xmath105 , and @xmath106 denote , respectively , the moment of inertia , internal thermal energy , internal kinetic energy , gravitational energy , the sum of the thermal surface pressure and dynamical surface pressure , internal magnetic energy , the magnetic surface pressure , and the maxwell stress - energy tensor .
the internal thermal , kinetic , and magnetic energies are always positive .
other terms can be either positive or negative .
in particular , the gravitational term can be positive in crowded environments where the background gravitational field dominates the core dynamics ( see e.g. , * ? ? ? * ) .
in fact , as shown below , several cores have positive gravitational terms in our simulations , although no cores have positive gravitational terms for the initially magnetically subcritical case where the turbulent motions are not so strong because of the effects of magnetic cushion ( see fig .
16 of * ? ? ?
the second term on the left - hand side of equation ( [ eq : virial ] ) denotes the time derivative of the flux of moment of inertia through the core boundary .
as is the standard practice @xcite , we ignore the left - hand side of equation ( [ eq : virial ] ) in our discussion and consider a core to be in virial equilibrium if the sum @xmath107 .
the equilibrium line is shown in figure [ fig : virial analysis ] , where the sum of the surface terms ( @xmath108 ) is plotted against the gravitational term ( @xmath102 ) , both normalized to the sum of the internal terms ( @xmath109 ) .
for the surface term , the external kinetic term is generally much larger than the magnetic term ( @xmath105 ) .
for all three models , the majority of the cores lie below the virial equilibrium line , and thus they are expected to be unstable to contraction .
in addition , the surface term appears more important than the gravitational term for most of the cores . in the absence of the magnetic field ,
about 10 @xmath110 20 % of the cores have the normalized surface term of @xmath111 , whereas no cores have the normalized surface term of @xmath112 in the presence of the moderately - strong magnetic field .
the magnetic field tends to reduce the contribution of the surface terms to the core dynamical state , although the surface term is still dominant even in the presence of the strong magnetic field .
the core virial state is in contrast to the initially - magnetically subcritical case . for the initially - magnetically subcritical case , for most of the cores , the surface term ( @xmath108 ) and gravitational term ( @xmath102 ) tend to be smaller than the sum of the internal terms , @xmath109 @xcite , and thus most of the cores are distributed around the virial equilibrium line .
in addition , more massive cores tend to be distributed below the virial equilibrium line , and the importance of the self - gravity tends to depend on the core mass .
the gravitational term tends to be more important for more massive cores for the initially magnetically subcritical case , implying that the gravitational collapse induced by ambipolar diffusion regulates the core evolution .
in contrast , for the magnetically - supercritical cases studied in this paper , most of the cores are distributed further from the virial equilibrium line . the surface term ( @xmath108 )
tends to be larger than the sum of the internal terms , @xmath109 for most of the cores .
the core virial state shows no clear dependence on the mass .
the surface term , which is typically dominated by the ambient supersonic turbulence , is still important even for massive cores .
the above result is different from that of @xcite .
they found , in their ideal mhd simulations , a larger contribution from the surface term to the virial equation for a stronger magnetic field , which is oppose to the trend we found .
this difference likely comes from the fact that they only followed the very early phase of cloud evolution , when the initial supersonic turbulence has decayed significantly but star formation has yet to set in .
most of our cores are formed at much later times , after the decayed turbulence has been replenished by outflow feedback .
given this difference , it is perhaps not too surprising that the surface terms are much larger in our models than in theirs .
this is particularly true for the weaker magnetic field case , because of a more rapid and violent star formation . in our simulations ,
most of the identified cores have the densities smaller than the critical density beyond which the jeans condition is violated .
for example , the jeans condition is satisfied for 89 % ( model n1 ) , 85 % ( model w1 ) , and 95 % ( model s1 ) of the cores presented in this subsection . for the stronger initial magnetic field ,
the mean density of the cores tend to be smaller due to the magnetic support , and therefore the jeans condition is satisfied for a larger number of cores .
we also performed a run with 512@xmath113 for model s1 , with the threshold density fixed , so that the jeans condition is always satisfied in the entire computation box . to compare directly with the results presented in this subsection , we first resized the data obtained from the @xmath114 run to the @xmath48 grid data , and then applied the clumpfind method to the regridded data .
we confirmed that the statistical properties of the identified cores are essentially the same as those presented in this subsection .
for example , for all the cores identified from the @xmath114 data , the nonthermal 3d velocity dispersions of the identified cores stay below 1 km s@xmath23 , and its mean is estimated to be 0.43 km s@xmath23 , comparable to that of the @xmath48 data , 0.46 km s@xmath23 . after exploring the effects of the magnetic field on the core properties
, we now turn to the outflow strength . in our model ,
the strength of the outflow feedback is specified by the dimensionless parameter , @xmath52 .
this parameter is highly uncertain because it is generally difficult to determine the physical properties of molecular outflows accurately from observational data . in this paper , we adopt @xmath57 as a fiducial value , following @xcite .
recently , nakamura et al . (
2011 ) discovered many molecular outflow lobes toward an extremely - young cluster - forming clump , serpens south , on the basis of co @xmath115 observations , and derived the physical properties of the identified molecular outflows .
they roughly estimated @xmath116 for these outflows , assuming that the outflow gas is optically - thin .
if a substantial amount of the outflow gas is optically - thick , the outflow parameter @xmath52 would be larger . in the following ,
we compare the core properties of three models with different outflow strengths : @xmath117 , 0.1 , and 0.4 .
the fiducial value for plasma @xmath30 of 0.2 is adopted for these models . in figs .
[ fig : outflow]a through [ fig : outflow]c , we compare the velocity dispersion - radius relations for the moderately - strong magnetic models with no outflow feedback ( @xmath117 ) , weak outflow @xmath118 , and strong outflow ( @xmath57 ) . in the model with no outflow feedback ( model s3 ) , the number of identified cores is smaller than in the other two models , and the mean core mass tends to be slightly larger
. this difference may be due to the fact that the ambient turbulent motions are weaker without any outflow feedback and thus the turbulent fragmentation does not form smaller - scale structures as efficiently .
figure [ fig : outflow ] indicates that the velocity dispersion - radius relation does not depend strongly on the strength of the outflow feedback . for all three models , there is no clear relationship between the velocity dispersion and the core radius .
thus , it appears difficult to discriminate the effects of protostellar outflows based on the velocity dispersion - radius relation alone . the surface term - gravitational term plots in figs .
[ fig : outflow2]a through [ fig : outflow2]c indicate that the dynamical states of the cores depend somewhat on the outflow strength . in the absence of the outflow feedback
, more cores are located near the dashed line ( where the two terms are equal ) .
the gravitational term appears more important for this model than for the other two , although the surface term is clearly still important .
this is due , at least in part , to the fact that the global gravitational infall can generate fast motions even in the absence of protostellar outflow driven turbulence .
it is difficult , however , to gauge the effects of the outflow feedback from the virial parameters of the cores , which are often used in observational studies of the molecular cloud cores in star - forming molecular clouds .
the difficulty is illustrated in figs .
[ fig : outflow3]a through [ fig : outflow3]c , where we show the virial parameter - mass relations for the three models with different outflow strengths . in all cases ,
the distributions of the virial parameters are similar , following a power - law of @xmath88 with rather small dispersions . in this subsection
, we investigate how well the outflow axes are aligned with the global or initial magnetic field direction in parsec - scale cluster forming clumps . in our simulations ,
the outflow direction is set to the local magnetic field direction at the position of each formed star .
this should be a reasonable approximation since the outflows are most likely driven along the local magnetic field by field line twisting due to disk rotation , ( e.g. , * ? ? ?
figure [ fig : outflow4 ] shows the histogram of the outflow direction as a function of the angle between the outflow axis and the initial magnetic field direction , for the model with a moderately - strong magnetic field ( model s1 ) .
all the outflows created by the time when @xmath119 are shown .
the outflow direction shows a broad distribution , taking its maximum at the mean angle of about 30@xmath120 .
the conclusion is that , although there is some preference for the outflows to align with the global magnetic field direction , the alignment is not as strong as one may naively expect .
this is because most of the dense cores are formed through outflow - induced turbulent compression , which can push the magnetic field on the core scale away from the global field direction . on smaller scales ,
the misalignment in field direction can be further amplified by gravitational collapse .
the lack of a complete alignment between the outflow axes and the global field direction is thus expected .
it can not be used to argue against the presence of a moderately - strong global magnetic field .
here , we compare the core properties obtained in section [ sec : results ] with observations .
we use the core data toward two nearby parsec - scale cluster - forming clumps : the @xmath25 ophiuchus main cloud and ngc 1333 in perseus . in the @xmath25 ophiuchus region
, @xcite identified 68 h@xmath26co@xmath62 ( @xmath27 ) cores by applying clumpfind to the 3d position - position - velocity cube data obtained using nobeyama 45 m telescope . in ngc 1333 , @xcite identified 93 n@xmath15h@xmath62 ( @xmath27 ) dense cores , again using clumpfind .
since both the h@xmath26co@xmath62 ( @xmath27 ) and n@xmath15h@xmath62 ( @xmath27 ) transitions have critical densities of order @xmath59 @xmath13 , the h@xmath26co@xmath62 and n@xmath15h@xmath62 cores should have densities comparable to our simulated cores , which were identified using a threshold density @xmath59 @xmath13 .
since the moderately - strong field model yields a core velocity dispersion that is in better general agreement with observations than the non - magnetic or weak - field model , we will concentrate on comparing its cores with the @xmath25 oph and ngc 1333 data . in fig .
[ fig : obs]a , we compare the 1d fwhm nonthermal velocity width - radius relations for the @xmath25 oph cores of @xcite and our simulated cores ( see section [ sec : results ] ) .
the 3d velocity dispersions discussed in section [ sec : results ] was converted into the 1d fwhm velocity widths using @xmath121 .
our simulated cores tend to be somewhat smaller than the @xmath25 oph cores ; the average radius of the @xmath25 oph cores is estimated at 0.04 pc , about 1.7 times that of model s1 . aside from the size difference , the observed and simulated data are almost indistinguishable .
neither shows a clear dependence of linewidth with size , except for a weak trend for the linewidth to increase slightly with radius .
neither follows the larson s linewidth - size relation .
the observations are also consistent with the cores formed from initially - magnetically subcritical clouds ( see fig
. 18 of * ? ? ?
we did not plot the ngc 1333 cores on fig .
[ fig : obs]a because the core radius in table 1 of @xcite was quoted in single significant digit , which artificially introduces a large scatter in the plot .
nevertheless , inspection by eye reveals no clear correlation between the velocity width and core radius , broadly consistent with the data for both the @xmath25 oph cores and the simulated cores of model s1 .
figure [ fig : obs]b compares the virial parameter - mass relations of the simulated and observed cores .
the red and blue dots , and black crosses denotes the @xmath25 oph @xcite and ngc 1333 cores @xcite , and the cores of model s1 , respectively . for the ngc 1333 cores , we recalculated the virial masses by including the thermal contribution from the gas with a temperature of 15 k and adopting a distance of 235 pc @xcite , instead of 300 pc used in @xcite .
we also recalculated the lte masses by adopting the n@xmath15h@xmath62 fractional abundance relative to h@xmath15 of @xmath122 derived from observations toward @xmath25 oph because @xcite determined their adopted n@xmath15h@xmath62 fractional abundance such that the points are distributed evenly about the line of equality ( @xmath123 ) in their @xmath124-@xmath125 plot .
this n@xmath15h@xmath62 fractional abundance is determined by taking average of the values obtained toward oph a @xcite and b2 subclumps @xcite .
figure [ fig : obs]b indicates that the virial parameter - core mass relation of the nearby cluster - forming clumps follows roughly a single power - law of @xmath126 with a small dispersion , which is again in good agreement with model s1 that has a moderately - strong magnetic field .
a dynamically important magnetic field may have played a role in core formation in these regions . besides the @xmath25 oph region
, @xcite found that the velocity dispersions of h@xmath26co@xmath62 cores are almost independent of the core radius in orion a as well , even though the latter is more than a factor of 3 farther away .
the coarser spatial resolution for orion means that its cores may contain ( smaller , @xmath25 oph - like ) cores blending together .
if this is true , then the nearly flat velocity - width - radius relation observed in both regions suggests that the inter - core motions among the neighboring cores are almost comparable to the internal motions in the individual cores .
such a feature was pointed out by @xcite who measured the velocity difference among neighboring cores using n@xmath15h@xmath62 ( @xmath27 ) .
if the velocity - width - radius relation is flat ( @xmath127 constant ) and the core mass is proportional to @xmath128 , then the virial parameter is scaled as @xmath129 , a similar power - law to that derived by @xcite .
this again suggests the importance of ambient turbulent pressure in dynamics of the cores as discussed in section [ sec : results ] . to isolate the effect of magnetic field on clustered star formation , we present in fig . [ fig : sfe]a the time evolution of star formation efficiency ( sfe ) for three models with different field strengths but no outflow feedback .
our simulations show that , as expected , the magnetic field tends to slow down star formation , especially at relatively late times .
the rate of star formation remains rather high , however , particularly at early times .
for example , from the formation of the first star to the time when the sfe reaches @xmath130 , the average star formation efficiency per global free - fall time ( sfr@xmath131 ; krumholz & mckee 2005 ) is estimated to be 29 % , 27 % , and 21 % , respectively , for models n2 , w2 , and s2 .
these values are larger than the observationally inferred values of a few % @xcite .
the large sfr@xmath131 in the absence of outflow feedback is in agreement with previous studies ( see also the appendix for the analytic formula of sfr@xmath131 ) .
for example , using mhd sph simulations , @xcite and @xcite followed the evolution of 50 @xmath28 clumps with @xmath132 until @xmath133 .
although their initial model clump mass is too small compared to nearby cluster - forming clumps like @xmath25 oph and ngc 1333 ( which have masses of order @xmath134 m@xmath2 ) , their star formation efficiencies per free - fall time are estimated to be over 10 % .
these simulations imply that it is difficult to reduce the star formation rate to the observed level by a moderately - strong magnetic field alone , and other factors are needed to significantly retard star formation in cluster - forming clumps .
one way to slow down star formation is the feedback from forming stars .
@xcite found that radiative feedback from forming stars does not change the global star formation efficiency in a parent clump much , although it significantly suppresses the small - scale fragmentation by increasing the temperature in the high - density material near the protostars .
figure [ fig : sfe]b shows that the star formation rate is greatly reduced by the inclusion of outflow feedback , even in the presence of a relatively weak magnetic field , in agreement with previous studies @xcite . the average star formation efficiency per global free - fall time is estimated to be 8 % , 6 % , and 4 % for model n1 ( no magnetic field ) , w1 ( weak field ) , and s1 ( moderately - strong field ) , respectively . for the model with the moderately - strong field , the sfr@xmath131 is more or less comparable to the observed value .
the reason for the large reduction in sfr@xmath131 can be seen in fig .
[ fig : magnetic field ] , where the magnetic energy is plotted against the evolution time , for models w1 and s1 . as shown in fig .
[ fig : magnetic field]a , for the moderately - strong field case , the total magnetic energy is dominated by the background uniform field that does not contribute to the force balance in the initial cloud .
we therefore illustrate in fig .
[ fig : magnetic field]b the time evolution of the magnetic energy stored in the distorted component that was amplified by supersonic turbulence . here , we computed the magnetic energy stored in the distorted component by subtracting the initial magnetic energy from the total magnetic energy . for comparison , we also plotted in fig .
[ fig : magnetic field]b the evolution of the total kinetic energy of the clump . since the kinetic energy is often dominated by the unbound high - velocity gas associated with the active outflows , we approximate the total kinetic energy of the clump as the kinetic energy of the gas whose velocity is smaller than 10 @xmath135 . for both models ,
the amplified components tend to increase with time . for the model with the moderately - strong field ,
the amplified component begins to oscillate around a level value after a free - fall time .
for both models , the magnetic energy of the amplified component becomes comparable to the kinetic energy of the dense gas by the end of the computation , indicating that a quasi - equipartition has been reached ( see also * ? ? ? * ) .
for the weak field case , the amplified component is more important than the initial uniform field , resulting in a significantly - distorted magnetic field structure .
such a highly - distorted magnetic field can be seen in the 3d bird s eye view of the density and magnetic field distributions of model w1 , which is presented in the left panel of fig .
[ fig:3d ] .
in contrast , the global magnetic field is well - ordered for the case of moderately - strong initial magnetic field , which is shown in the right panel of fig .
[ fig:3d ] .
polarization maps of submillimeter thermal dust emission have recently been obtained for nearby star forming regions ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . here , we present the polarization maps derived from the simulation data for the two magnetically - supercritical models with different initial magnetic field strengths ( models w1 and s1 which have the magnetic flux - to - mass ratios of @xmath136 and 1.4 , respectively ) .
we computed the polarized thermal dust emission from the mhd model following @xcite .
we neglect the effect of self - absorption and scattering because we are interested in the thermal dust emission at submillimeter wavelengths .
we further assume that the grain properties are constant and the temperature is uniform .
the polarization degree is set such that the maximum is equal to 15 % .
figure [ fig : polarization ] shows the dust polarization maps calculated from models w1 and s1 .
only a small portion of the computation box is shown in each panel of fig .
[ fig : polarization ] . as expected from fig .
[ fig:3d ] , in the presence of a weak magnetic field , the spatial distribution of the polarization vectors has relatively large fluctuations .
the polarization degree tends to be smaller in several parts where the magnetic fields are strongly distorted . in the densest parts , the polarization vectors appear more or less parallel to the local elongated structures or dense filaments , whereas there is no clear correlation between the vectors and density distribution in less dense parts .
this may be due to the fact that the local dense filaments are created by turbulent compression that preferentially enhances the magnetic field component transverse to the shock plane .
in contrast , in the presence of the strong magnetic field , the filamentary structure is prominent and the filament axes tend to be perpendicular to the polarization vectors that are almost parallel to the initial magnetic field direction , suggesting that the gas flow is channeled preferentially by the strong magnetic field to form the filaments .
the polarization observations can thus provide a handle on the magnetic field strength of cluster - forming clumps .
recent polarization observations of cluster - forming clumps show that the global magnetic field lines are more or less spatially well - ordered ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
for example , @xcite found that the serpens cloud core is threaded by a hour - glass shaped well - ordered magnetic field and is elongated in the cross - field direction .
@xcite found that the serpens south filamentary infrared dark cloud appears to be threaded by more or less straight global magnetic field .
the observed spatially - ordered magnetic field structures imply that the magnetic fields in the nearby cluster - forming clumps are likely to be at least moderately - strong and dynamically important . in figs .
[ fig : cmf]a , [ fig : cmf]b , and [ fig : cmf]c , we plot the core mass functions ( cmfs ) for our identified cores for models n1 , w1 , and s1 , respectively . to enlarge the number of sample cores , we added up all the cores identified at the three stages when sfe has reached 0.08 , 0.12 , and 0.16 .
in addition , we added up the cores identified from the runs using the different initial turbulent realization for which the different random numbers were used for both the amplitude and phase .
the total number of cores so identified is over 500 for each model .
the overall shapes of the cmfs are similar in all three cases , suggesting the shape of the cmf is relatively insensitive to the initial magnetic field strength .
this is different from the turbulent fragmentation scenario proposed by @xcite where a weak magnetic field is required to reproduce the cmf that is similar in shape to the salpeter imf ( see also * ? ? ?
in our simulations , a moderately - strong magnetic field can produce a cmf that resembles the salpeter imf as well .
we note two interesting features of the computed cmfs .
first , there appear to be a lack of massive cores compared to the salpeter stellar imf .
this is not necessarily a problem , because it is unclear whether the nearby cluster - forming clumps such as @xmath25 oph and ngc 1333 that we aim to simulate would ever produce massive stars . alternatively
, massive stars can grow from initially less massive cores , fed by global collapsing flows towards the bottom of the clump potential ( smith et al .
2008 ; wang et al .
second , the turnover at the low mass end is less clear in the cmfs than in the kroupa or chabrier imf .
this may be because , at the low mass end , most of the cores are gravitationally unbound , and may not form stars .
further studies are required to determine the relation between the cmf and imf .
we have performed a set of 3d mhd simulations of cluster formation taking into account the effects of protostellar outflow feedback , and identified dense cores by applying a clumpfind algorithm to the simulated 3d density data cubes .
the main results are as follows .
dense cores do not follow larson s linewidth - size relation .
we find that the velocity dispersions of dense cores show little correlation with core radius , irrespective of the strength of the magnetic field and outflow feedback . in the absence of a magnetic field ,
the majority of the cores have supersonic velocity dispersions , whereas in the presence of a moderately - strong magnetic field , the cores tend to be subsonic or at most transonic .
we find that most of the cores are out of virial equilibrium , with the external pressure due to ambient turbulence dominating the self - gravity .
the core formation and evolution is largely controlled by the dynamical compression due to outflow - driven turbulence . such a situation is contrast to the strongly - magnetized ( magnetically subcritical ) case , where the self - gravity plays a more important role in the core dynamics , particularly for massive cores .
even an initially - weak magnetic field can retard star formation significantly , because the field is amplified by supersonic turbulence to an equipartition strength .
in such an initially weak field , the distorted field component dominates the uniform one .
in contrast , for a moderately - strong field , the uniform component remains dominant .
such a difference in the magnetic structure can be observed in simulated polarization maps of dust thermal emission .
recent polarization measurements show that the field lines in nearby cluster - forming clumps are spatially well - ordered , indicative of a moderately - strong , dynamically - important , field .
this work was supported in part by a grant - in - aid for scientific research of japan ( 20540228 and 22340040 ) and nasa ( nng06gj33 g and nnx10ah30 g ) .
the numerical calculations were carried out mainly on nec sx9 at the center for computational astrophysics ( cfca ) at national astronomical observatory of japan and on nec sx8 at yitp in kyoto university .
99 andr , p. , belloche , a. , motte , f. , & peretto , n. 2007 , , 472 , 519 arce , h. g. , borkin , m. a. , goodman , a. a. , pineda , j. e. , & halle , m. w. 2010 , apj , in press ballesteros - paredes , j. , gomez , g. c. , pichardo , b. , & vzquez - semadeni , e. 2009 , , 393 , 1563 bertoldi , f. & mckee , c. f. 1992 , , 395 , 140 bontemps , s. 2001 , , 372 , 173 butler , m. j. , & tan , j. c. 2009 , , 696 , 484 carroll , j. j. , frank , a. , blackman , e. g. , cunningham , a. j. , & quillen , a. c. 2009 , , 695 , 1376 crutcher , r. m. , 1999 , , 520 , 706 davis , c. j. , chrysostomou , a. , matthews , h. e. , jenness , t. , & ray , t. p. 2000 , , 530
, l115 m. 2007 , , 61 , 262 dib , s. , kim , j. , vzquez - semadeni , e. , burkert , a. , & ahadmehri , m. 2007 , , 61 , 262 di francesco , j. , andre , p. , & myers , p. c. 2004 , , 617 , 425 elmegreen , b. g. 2000 , , 530 , 277 federrath , c. , sur , s. , schleicher , d. r. g. , banerjee , r. , & klessen , r. s. 2011 , in press ( axxiv:1102.0266 ) .
friesen , r. k. , di francesco , j. , shimajiri , y. , & takakuwa , s. , 2010 , , 708 , 1002 girart , j. m. , beltrn , m. t. , zhang , q. , rao , r. , & estalella , r. 2009 , science , 324 , 1408 hartmann , l. w. , ballesteros - paredes , j. , bergin , e. a. 2001 , , 562 , 852 hennebelle , p. , whitworth , a. p. , gladwin , p. p. , & andr , p. 2003 , , 340 , 870 higuchi , a. e. , kurono , y. , saito , m. , & kawabe , r. 2009 , , 705 , 468 hirota , t. et al .
2008 , , 60 , 37 houde , m. et al .
2004 , 604 , 717 johnstone , d. , wilson , c. d. , moriarty - schieven , g. , et al .
2000 , , 545 , 327 krumholz , m. r. & mckee , c. f. 2005 , , 630 , 250 klessen , r. , heitsch , f. , & mac low , m .-
2000 , , 535 , 887 krumholz , m. r. & tan , j. c. , 2007 , , 656 , 959 larson , r. b. 1981 , , 194 , 809 loinard , l. , torres , r. m. , mioduszewski , a. j. & rodriguez , l. f. 2008 , , 675 , l29 lombardi , m. , lada , c. j. & alves , j. 2008 , , 480 , 785 li , z .- y .
& nakamura , f. 2006 , , 640 , l187 li , z .- y . ,
wang , p. , abel , t. , & nakamura , f. 2010 , , 720 , 26 loren , r. t. 1989 , , 338 , 902 mac low , m .-
m . , klessen , r. s. , burkert , a. , & smith , m. d. 1998 , , 80 , 2754 mac low , m .-
1999 , , 524 , 169 maruta , h. , nakamura , f. , nishi , r. , ikeda , n. , kitamura , y. 2010 , , 714 , 680 matsumoto , t. , nakazato , t. , & tomisaka , k. 2006 , , 637 , l105 matthews , b. c. , mcphee , c. a. , fissel , l. m. , & curran , r. l. 2009 , , 182 , 143 matzner , c. d. , & mckee , c. f. 2000 , , 545 , 364 matzner , c. d. 2007 , , 659 , 1394 maury a. , andr , p. , & li , z .- y . , 2009 , 499 , 175 mckee , c. f. , & zweibel , e. g. 1992 , , 399 , 551 motte , f. , andr , p. , & neri , r. 1998 , , 336 , 150 myers , p. c. 2009 , , 700 , 1609 nakamura , f. & li , z .- y . , 2007 , 662 , 395 nakamura , f. & li , z .- y . , 2008 , 687 , 354 nakamura , f. et al .
2011a , , 726 , 46 nakamura , f. et al . 2011b , , in press ( arxiv:1105.4481 ) nakano , t. , & nakamura , t. 1978 , , 30 , 681 ostriker , e. c. , stone , j. m. , & gammie , c. f. , 2001 , , 546 , 980 padoan , p. , goodman , a. , draine , b. t. , juvela , m. , nordlund , a. , & rognvaldsson , o. e. 2001 , , 559 , 1005 padoan , p. & nordlund , a. 2002 , , 576 , 870 pelkonen , v .- m .
, juvela , m. , & padoan , p. 2007
, , 461 , 551 price , d. j. , & bate , m. r. 2008 , , 385 , 1820 price , d. j. , & bate , m. r. 2009 , , 398 , 33 rathborne , j. m. , jackson , j. m. , & simon , r. 2006 , , 641 , 389 ridge , n. a. , wilson , t. l. , megeath , s. t. , allen , l. e. , & myers , p. c. 2003 , , 126 , 286 smith , r. j. , clark , p. c. , & bonnell , i. a. 2008 , , 391 , 1091 stanke , t. , smith , m. d. , gredel , r. , & khanzadyan , t. 2006 , , 447 , 609 sugitani , k. , nakamura , f. , tamura , m. , watanabe , m. , kandori , r. , nishiyama , s. , kusakabe , n. , hashimoto , j. , nagata , t. , & sato , s. 2010 , , 716 , 299 sugitani , k. et al .
2011 , , 734 , 63 testi , l. & sergent , a. 1998 , , 508 , l91 tilley , d. , & pudritz , r. 2007 , , 382 , 73 walsh , a. j. , myers , p. c. , di francesco , j. , mohanty , s. , bourke , t. l. , gutermuth , r. & wilner , d. 2007 , , 655 , 958 wang , p. , li , z .- y . , abel , t. , & nakamura , f. 2010 , , 709 , 27 williams , j. p. , de geus , e. j. , & blitz , l. 1994 , , 428 , 693
using 3d mhd numerical simulations of cluster formation , @xcite and @xcite demonstrated that the protostellar outflow - driven turbulence can keep a pc - scale , cluster - forming clump close to a virial equilibrium long after the initial turbulence has decayed away . here
, we derive an analytic formula of star formation rate in a cluster - forming clump that keeps its virial equilibrium by the protostellar outflow feedback .
numerical simulations of protostellar turbulence indicate that the dissipation rate of the turbulence momentum , @xmath137 , balances the momentum injection rate by the protostellar outflow feedback , @xmath138 , so that the clump can be kept close to a virial equilibrium , @xmath139 the dissipation rate of the turbulence momentum , @xmath137 , can be written as @xmath140 where @xmath141 @xcite . the virial velocity , @xmath142 , is given by @xmath143 the momentum dissipation time is @xmath144 .
the dimensionless parameter of order unity , @xmath145 , measures the effects of a nonuniform or nonspherical mass distribution . for a uniform sphere and a centrally - condense sphere with @xmath85 , @xmath146 and 5/3 , respectively . in the following analysis
, we adopt @xmath147 because the cluster - forming clumps tend to be centrally - condensed .
we take into account the magnetic support by multiplying the virial velocity by a factor @xmath148 , where @xmath149 .
the momentum injection rate by the protostellar outflows is given by @xmath150 where @xmath54 is the flow speed and @xmath151 is the star formation rate . using equations ( [ eq : turbulence ] ) and ( [ eq : outflow ] ) , the star formation rate is rewritten as @xmath152
if we normalize @xmath153 to @xmath154 , then the star formation efficiency per free - fall time is given by @xmath155 where the surface density is @xmath156 and the free - fall time is @xmath157 .
figure [ fig : sfr ] shows the dependence of sfr@xmath131 on the mass and radius , obtained from the above equation .
the crosses and diamonds indicate the cluster - forming clumps observed in the c@xmath158o ( @xmath27 ) line by @xcite and @xcite , respectively .
our model suggests that the star formation rate per free - fall time ranges from 1 % to 5 % for the observed cluster - forming clumps in the solar neighborhood when the protostellar outflow feedback maintains the supersonic turbulence in the clumps , indicating that it takes about ( 2 @xmath110 10 ) @xmath159 for the star formation efficiency to reach about ( 10 @xmath110 20 ) % .
llll n1 & @xmath160 & 0.4 & no b field + n2 & @xmath160 & 0.0 & no b field , no feedback + w1 & 2 & 0.4 & weak b field + w2 & 2 & 0.0 & weak b field , no feedback + s1 & 0.2 & 0.4 & moderately - strong b field + s2 & 0.2 & 0.0 & moderately - strong b field , no feedback + s3 & 0.2 & 0.1 & moderately - strong b field , weak feedback [ tab : model ] lllll model n1 & & & & + @xmath70 ( pc ) & 0.013 & 0.041 & 0.020 @xmath161 0.005 & 0.020 + @xmath162 ( m@xmath2 ) & 0.075 & 3.51 & 0.67 @xmath161 0.67 & 0.40 + @xmath163 ( km s@xmath23 ) & 0.13 & 4.53 & 1.06 @xmath161 0.77 & 0.86 + model w1 & & & & + @xmath70 ( pc ) & 0.013 & 0.034 & 0.020 @xmath161 0.006 & 0.019 + @xmath162 ( m@xmath2 ) & 0.076 & 3.90 & 0.62 @xmath161 0.77 & 0.31 + @xmath163 ( km s@xmath23 ) & 0.13 & 14.4 & 0.99 @xmath161 1.66 & 0.47 + model s1 & & & & + @xmath70 ( pc ) & 0.013 & 0.048 & 0.024 @xmath161 0.008 & 0.023 + @xmath162 ( m@xmath2 ) & 0.073 & 4.39 & 0.86 @xmath161 1.03 & 0.47 + @xmath163 ( km s@xmath23 ) & 0.086 & 2.64 & 0.46 @xmath161 0.52 & 0.31 | we investigate the physical properties of dense cores formed in turbulent , magnetized , parsec - scale clumps of molecular clouds , using three - dimensional numerical simulations that include protostellar outflow feedback .
the dense cores are identified in the simulated density data cube through a clumpfind algorithm .
we find that the core velocity dispersion does not show any clear dependence on the core size , in contrast to larson s linewidth - size relation , but consistent with recent observations . in the absence of a magnetic field ,
the majority of the cores have supersonic velocity dispersions .
a moderately - strong magnetic field reduces the dispersion to a subsonic or at most transonic value typically .
most of the cores are out of virial equilibrium , with the external pressure dominating the self - gravity .
the implication is that the core evolution is largely controlled by the outflow - driven turbulence .
even an initially - weak magnetic field can retard star formation significantly , because the field is amplified by the outflow - driven turbulence to an equipartition strength , with the distorted field component dominating the uniform one .
in contrast , for a moderately - strong field , the uniform component remains dominant .
such a difference in the magnetic structure is evident in our simulated polarization maps of dust thermal emission ; it provides a handle on the field strength .
recent polarization measurements show that the field lines in cluster - forming clumps are spatially well - ordered .
it is indicative of a moderately - strong , dynamically important , field which , in combination with outflow feedback , can keep the rate of star formation in embedded clusters at the observationally - inferred , relatively - slow rate of several percent per free - fall time . |
the point group symmetry of graphene on an incommensurate substrate is given by the intersection of the point group of graphene , @xmath81 , with that of its substrate .
for a perfectly aligned ( @xmath82 ) inversion - symmetric substrate , with either a single ( dominant ) atom per unit cell or two identical atoms arranged in a honeycomb lattice , the point group symmetries of the substrate and graphene coincide .
the corresponding hamiltonian , eq . , with moir harmonics orientated as per fig .
[ fig : fig1 ] ( a ) , must necessarily commute with the operators corresponding to the elements of @xmath81 : @xmath83 , @xmath84 and @xmath85 which describe @xmath86 rotations and reflections that either exchange or preserve the graphene sublattices .
the operators for @xmath83 and @xmath85 involve the valley exchanging matrices @xmath87 resulting in that the symmetry of the hamiltonian restricted to the @xmath16 valley , as well as the @xmath16 valley bandstructure , is reduced to @xmath88 , where @xmath89 has no intervalley structure .
each of the @xmath90 terms are odd under @xmath83 , while the @xmath91 and @xmath92 terms are odd under @xmath85 , so that these terms are forbidden for the perfectly aligned inversion - symmetric system described above .
the point group of substrates with the honeycomb lattice and two non - equivalent atoms per unit cell , such as hbn , only possesses the @xmath93 and @xmath85 symmetries which allow inversion - asymmetric parameters @xmath94 to take a finite value . for a finite misalignment angle , the reflection symmetries of graphene and the substrate do not coincide , and the moir harmonics become misaligned , by an angle @xmath95 , from those in fig .
[ fig : fig1 ] ( a ) .
however , the moir harmonics may be brought back into alignment using the transformation @xmath96 , and the @xmath91 and @xmath92 terms , which are no longer forbidden , may be gauged away .
this procedure restores the reflection symmetries to the hamiltonian , despite their absence in the geometry of the moir pattern for finite misalignment angle .
the symmetries described above can be used to gain a deeper understanding of the mdps discussed in the main text .
the @xmath16 valley plane wave states from the three equivalent sbz corners , @xmath97 , which form the basis for @xmath98 , eq . , transform into each other on application of symmetry operators of @xmath40 . in the same basis ,
the symmetry operators acting on @xmath99 take the form of matrices @xmath100 for the inversion - symmetric superlattice perturbation , the singlet eigenstate of @xmath98 is given by @xmath101 .
the action of matrices from eq . on this state show that it transforms according to the one - dimensional irreducible representations of @xmath40 : either @xmath75 for @xmath102 or @xmath76 for @xmath103 , indicating evenness or oddness under @xmath84 respectively .
similarly , the doublet states of @xmath98 , @xmath104 and @xmath105 transform as the two - dimensional irreducible representation , @xmath74 , and their degeneracy is therefore protected by the @xmath40 symmetry .
the three anisotropic mdps can be understood using the compatibility relations in the group appropriate for the sbz edge , @xmath106 .
this group only supports one - dimensional irreducible representations @xmath75 and @xmath76 with the doublet states reducing as @xmath107 . for a given band , @xmath108 ,
the split bands at @xmath61 and @xmath60 belong to different irreducible representations of @xmath109 and therefore can not be joined along the sbz edge . instead , if both of these bands are closer to zero energy than the doublet states , they must each be joined to one of the doublet bands at the opposite sbz corner . thus , along the sbz edge , a crossing of the split bands is required resulting in the mdps illustrated in the valence band for the top row of fig .
[ fig : fig2 ] .
the point charge model analysed in this appendix mimics the effect of the quadrupole electric moment of the atoms in the top layer of the substrate . in application to the graphene - hbn system
, we neglect the potentials of the quadrupole moments of the boron atom , which have only @xmath110-orbitals occupied by electrons , and replace nitrogen sites by a point core charge @xmath111 compensated by the spread out cloud of the @xmath112-electrons , which we replace by a homogeneous background charge density , giving @xmath113 per hexagonal unit cell of the substrate .
this model gives an example of an inversion - symmetric moir superlattice .
the matrix elements of the resulting perturbation , taken between sublattice bloch states @xmath114 and @xmath115 ( @xmath116 ) , acting on the low energy dirac spinors of the graphene @xmath16 valley , are given by the long wavelength components of in eq .
@xmath118 are positions of nitrogen sites and @xmath119 is the total area of the graphene sheet ; @xmath120 are bloch wavefunctions of graphene @xmath112-electrons exactly at the @xmath16 point .
then the fourier transform has been used to write @xmath121 in terms of a sum over substrate reciprocal lattice vectors , @xmath122 , and graphene reciprocal lattice vectors , @xmath123 and @xmath124 .
nearest neighbour vectors , @xmath125 , are @xmath126 and @xmath127 , so that @xmath128 .
the homogeneous background charge has not been included in eq . since
its only role is to exclude @xmath129 from the sum .
the long wavelength terms in the first exponential of the second line of eq . determine @xmath130 .
the dimensionless integral , @xmath131 is written in terms of the fourier transform of the hydrogen - like graphene @xmath132 orbitals with an effective bohr radius @xmath133 , @xmath134 the integral , @xmath135 , rapidly decays as a function of the magnitude all its arguments so that we limit the sum in eq .
to only several terms such that @xmath136 , with @xmath137 where @xmath138 . for the graphene layer , @xmath139 and for the graphene - hbn heterostructure @xmath140 .
the carbon p@xmath3 orbitals may have a different effective bohr radius compared to hydrogen .
the range of values quoted for @xmath141 in table [ table_1 ] corresponds to the interval @xmath142 , indicated by the black double - arrow in fig .
[ fig : fig4 ] .
interlayer separation @xmath143 is taken from ref .
@xcite . both the dominance of the simplest moir harmonics and the finite values for the off - diagonal terms @xmath144 and @xmath91 stem from the three dimensional treatment of the substrate potential .
the potential is strongest near the substrate and therefore a greater proportion of the integral @xmath135 comes from the region near the substrate , where the graphene p@xmath3 orbitals are broad and therefore have both rapidly decaying fourier components and significant overlap with their neighbours .
this contrasts with the model employed in ref .
@xcite which is based on a two - dimensional substrate potential resulting in @xmath145 .
, as a function of the effective bohr radius of the graphene p@xmath34 orbitals , for various choices of interlayer separation @xmath0 . to demonstrate convergence of the sum in eq .
, dashed lines show @xmath146 for the same values of @xmath0.,scaledwidth=45.0% ] in ref .
@xcite , kindermann _
et al . _
modelled a hbn substrate as a lattice of p@xmath147 orbitals onto which the graphene electrons can hop .
this treatment , extended from a model of twisted bilayer graphene @xcite , assumed equal values for the hopping integral to the boron and nitrogen sites , with the difference between the two sublattices arising from their different on - site energies .
here we consider an inversion - symmetric version of the hopping model of ref .
@xcite , assuming that coupling between graphene and the hbn layer is dominated by the hopping to only one of the two sublattices ( e.g. boron ) . using @xmath148 theory , this coupling can be written as @xcite @xmath149 neglecting a non - oscillatory term , which corresponds to a trivial constant energy shift , eq . as applied to graphene electrons in valley @xmath16 ,
leads to the moir hamiltonian , eq . , with @xmath150 the parameters of the superlattice perturbation given in table [ table_1 ] of the main text , correspond to @xmath151 , @xmath152 and @xmath153 , in accordance with ref .
@xcite . for the perfectly aligned system
, we always find @xmath154 , which is a consequence of the reflection symmetries present in the perfectly aligned substrate - graphene system ( see appendix 1 ) . | using a general symmetry - based approach , we provide a classification of generic miniband structures for electrons in graphene placed on substrates with the hexagonal bravais symmetry .
in particular , we identify conditions at which the first moir miniband is separated from the rest of the spectrum by either one or a group of three isolated mini dirac points and is not obscured by dispersion surfaces coming from other minibands . in such cases the hall coefficient exhibits two distinct alternations of its sign as a function of charge carrier density .
recently , it has been demonstrated that the electronic quality of graphene - based devices can be dramatically improved by placing graphene on an atomically flat crystal surface , such as hexagonal boron nitride ( hbn ) @xcite . at the same time , graphene s electronic spectrum also becomes modified , acquiring a complex , energy - dependent form caused by incommensurability between the graphene and substrate crystal lattices @xcite . in particular , for graphene placed on hbn , the difference between their lattice constants and crystallographic misalignment generate a hexagonal periodic structure known as a moir pattern @xcite .
the resulting periodic perturbation , usually referred to as a superlattice , acts on graphene s charge carriers and leads to multiple minibands and the generation of secondary dirac - like spectra .
the resulting new dirac fermions present yet another case where graphene allows mimicking of qed phenomena under conditions that can not be achieved in particle physics experiments .
in contrast to relativistic particles in free space , the properties of secondary dirac fermions in graphene can be affected by a periodic sublattice symmetry breaking and modulation of carbon - carbon hopping amplitudes , in addition to a simple potential modulation . the combination of different features in the modulation results in a multiplicity of possible outcomes for the moir miniband spectrum in graphene which we systematically investigate in this article . to describe the effect of a substrate on electrons in graphene at a distance , @xmath0 , much larger than the spacing , @xmath1 , between carbon atoms in graphene s honeycomb lattice
, we use the earlier observation @xcite that , at @xmath2 the lateral variation of the wavefunctions of the p@xmath3 carbon orbitals is smooth on the scale of @xmath1 .
this is manifested in the comparable sizes of the skew and vertical hopping in graphite and permits an elegant continuum - model description @xcite of the interlayer coupling in twisted bilayers and the resulting band structure .
a similar idea applied to graphene on a hbn substrate @xcite suggests that a substrate perturbation for dirac electrons in graphene can be described in terms of simple harmonic functions corresponding to the six smallest reciprocal lattice vectors of the moir superlattice .
below , we shall use a similar approach to analyse the generic properties of moir minibands for electrons in graphene subjected to a substrate with a hexagonal bravais lattice with a slightly different lattice constant of @xmath4 , @xmath5 , compared to that of @xmath6 for graphene , and a small misalignment angle , @xmath7 .
the moir pattern harmonics are described by vectors @xmath8\ ! \left(\!0,\!\frac{4
\pi}{3 a}\!\right)\!,\end{aligned}\ ] ] with length @xmath9 , which can be obtained from each other by the anticlockwise rotation , @xmath10 . for a substrate with a simple hexagonal lattice or a honeycomb lattice with two identical atoms ,
the perturbation created for graphene electrons is inversion - symmetric . for a honeycomb substrate where one of the atoms would affect graphene electrons stronger than the other ( e.g. such as hbn , for which the occupancy and size of the p@xmath3 orbitals are different ) the moir potential can be modelled as a combination of a dominant inversion - symmetric part with the addition of a small inversion - asymmetric perturbation , @xmath11\cdot{\boldsymbol}\sigma\tau_3 + u_2 v \nabla f_2({\boldsymbol}r)\cdot{\boldsymbol}\sigma\tau_3 \label{eq : h } \\ & + \tilde u_0 v b f_2({\boldsymbol}r)+\tilde u_3 v b f_1({\boldsymbol}r)\sigma_3\tau_3 + \tilde u_1 v \left[{\boldsymbol}l_z\times\nabla f_1({\boldsymbol}r)\right]\cdot{\boldsymbol}\sigma\tau_3+\tilde u_2 v \nabla f_1({\boldsymbol}r)\cdot{\boldsymbol}\sigma\tau_3 .
\nonumber \end{aligned}\ ] ] the hamiltonian , @xmath12 , acts on four - component wavefunctions , @xmath13 , describing the electron amplitudes on graphene sublattices @xmath14 and @xmath15 and in two principal valleys , @xmath16 and @xmath17 .
it is written in terms of direct products @xmath18 , of pauli matrices @xmath19 and @xmath20 separately acting on sublattice and valley indices .
the first term in @xmath12 is the dirac part , with @xmath21 describing the momentum relative to the centre of the corresponding valley , with @xmath22 .
the rest of the first line in eq . describes the inversion - symmetric part of the moir perturbation , whereas the second line takes into account its inversion - asymmetric part . in the first line , the first term , with @xmath23 ,
describes a simple potential modulation . the second term , with @xmath24 , accounts for the @xmath14-@xmath15 sublattice asymmetry , locally imposed by the substrate .
the third term , with unit vector @xmath25 , describes the influence of the substrate on the @xmath14-@xmath15 hopping : consequently @xcite , this term can be associated with a pseudo - magnetic field , @xmath26 , which has opposite signs in valleys @xmath16 and @xmath17 .
each of the coefficients @xmath27 in eq . is a dimensionless phenomenological parameter with the energy scale set by @xmath28 , where @xmath29ev is the nearest neighbour hopping integral in the slonczewski - weiss tight binding model @xcite . concerning the inversion - asymmetric part , the second line in eq .
, we assume that @xmath30 .
note that the last term in each line can be gauged away using @xmath31 .
hamiltonian , @xmath12 , may be used to parametrise any microscopic model compatible with the symmetries of the system ( see appendix 1 ) and the dominance of the simplest moir harmonics , @xmath32 , in the superlattice perturbation .
the values that parameters @xmath33 take are listed in table [ table_1 ] for several models of graphene on an hbn substrate , both taken from the recent literature @xcite and analysed in appendix 2 , including a simple model in which the hbn substrate is treated as a lattice of positively charged nitrogen nuclei with a compensating homogeneous background of electron p@xmath34 orbitals .
the examples of model - dependent values of parameters @xmath33 , listed in table [ table_1 ] , indicate that the combination of several factors can strongly shift the resulting moir perturbation across the parameter space in fig .
[ fig : fig1 ] . that is why , in this work , we analyse the generic features of the miniband spectra generated by the moir superlattice , rather than attempt to make a brave prediction about its exact form for a particular substrate .
.the inversion - symmetric parameters , @xmath35 , for various models of the moir superlattice . in the 2d
charge modulation model @xcite , @xmath36 is a phenomenological parameter .
the g - hbn hopping model in ref .
@xcite used the hopping parameter from twisted bilayer graphene .
estimates in appendix 2 show that the sets of parameters using a model of point charges attributed to nitrogen sites and for the g - hbn hopping model are very similar . [ cols="<,^,^,^,^ " , ] in the absence of a magnetic field , the hamiltonian eq .
obeys time - reversal symmetry , which follows from both @xmath19 and @xmath37 changing sign upon the transformation @xmath38 @xcite . as a result , @xmath39
and we limit the discussion of minibands to the @xmath16 valley .
subject to this limitation the bandstructure for the inversion - symmetric superlattice perturbation obeys the @xmath40 symmetry .
moreover , using the commutation properties of @xmath19 one can establish that @xmath41 to calculate the miniband spectrum for @xmath12 in eq .
we perform zone folding ( in the graphene @xmath16 valley ) bringing states with momenta related by the reciprocal lattice vectors @xmath42 of the moir pattern to the same point of the superlattice brillouin zone ( sbz ) in fig .
[ fig : fig1](a ) .
then , we calculate the matrix elements of @xmath12 between those states and diagonalise the corresponding heisenberg matrix numerically exploring the parametric space @xmath43 of the dominant inversion - symmetric part of the moir perturbation shown in fig [ fig : fig1 ] ( b , c ) .
the size of the matrix is chosen to guarantee the convergence of the calculated energies for the three lowest minibands in both the conduction band ( @xmath44 ) and the valence band ( @xmath45 ) .
below , we discuss the generic features of the moir miniband spectra for the characteristic points in the parametric space @xmath43 , marked using black dots in fig .
[ fig : fig1](b , c ) , using both the numerically calculated dispersion surfaces in fig .
[ fig : fig2 ] and analytical perturbation theory analysis .
for the zero - energy dirac point in graphene , there are only the original @xmath46 states in each valley that appear at @xmath47 upon zone folding .
for all three characteristic spectra shown in fig .
[ fig : fig2 ] , for the inversion - symmetric moir perturbation , the gapless dirac spectrum persists at low energies near the conduction - valence band edge with almost unchanged dirac velocity , @xmath48v$ ] .
the inversion - asymmetric terms @xmath49 are able @xcite to open a minigap at the dirac point . for the point @xmath50 on the edge of the first sbz
, zone folding brings together two degenerate plane wave states , @xmath51 and @xmath52 .
the splitting of these degenerate states by the moir potential in eq . can be studied using degenerate perturbation theory .
the corresponding @xmath53 matrix , expanded in small deviation @xmath54 of the electron momentum from each of the three sbz @xmath55-points @xcite has the form @xmath56 for the inversion - symmetric perturbation , the dispersion relation resulting from eq .
contains an anisotropic mini dirac point ( mdp ) @xcite with dirac velocity component @xmath57 in the direction of the sbz edge and @xmath58 in the perpendicular direction .
this feature is clearly seen at the @xmath55-point of the first moir miniband in the valence band , in the top row of fig . [
fig : fig2 ] . note that the electron spectrum is not symmetric between the valence and conduction bands and that the mdps at the @xmath55-point in the conduction band are obscured by an overlapping spectral branch .
moving in parameter space , e.g. , along the line shown in fig .
[ fig : fig1](b ) , the positions of the three anisotropic mdps shift along the sbz edge towards the sbz corners : either @xmath59 , or @xmath60 , as shown by arrowed lines in fig .
[ fig : fig1](a ) . in general , a spectrum with three isolated mdps at the sbz edge is typical for the green volume in the parameter space in fig .
[ fig : fig1](b ) for the valence band , or fig .
[ fig : fig1](c ) for the conduction band .
in contrast , for @xmath43 in the clear part of the parameter space , mdps on the edge of the first sbz are overshadowed by an overlapping spectral branch , as is the case on the conduction band side for all three cases shown in fig .
[ fig : fig2 ] . for the points in fig .
[ fig : fig1](b , c ) on the red and blue surfaces , the three mdps reach the @xmath61-point , forming a triple degenerate band crossing , as in the valence band spectrum shown in the middle row of fig .
[ fig : fig2 ] , which can be traced using the perturbation theory analysis of the band crossing at @xmath61 discussed below . the third line in fig .
[ fig : fig2 ] shows the third type of spectrum of moir minibands , characteristic for the red and blue volumes of the parameter space in fig .
[ fig : fig1 ] .
the characteristic feature of such spectra consists in a single isolated mdp , at the @xmath62-point , in the valence band ( fig . [
fig : fig1](b ) ) or the conduction band ( fig .
[ fig : fig1](c ) ) . for the @xmath61 and @xmath60-points ,
zone folding brings together three degenerate plane wave states , @xmath63 , @xmath64 , and @xmath65 ( where @xmath66 ) , whose splitting is determined by @xmath67\!.\!\label{eq : kappa } \end{aligned}\ ] ] for @xmath68 , the inversion - symmetric terms in @xmath69 partially lift the @xmath70-point degeneracy into a singlet with energy @xmath71 and a doublet with energies @xmath72 , so that a distinctive mdp @xcite characterised by dirac velocity @xmath73\frac{v}{2}$ ] @xcite is always present at @xmath62 somewhere in the spectrum @xcite .
this behaviour reflects the generic properties of the symmetry group of wave vector @xmath61 which has the two - dimensional irreducible representation @xmath74 ( corresponding to the mdp ) and one - dimensional irreducible representations @xmath75 and @xmath76 .
note that each isolated mdp is surrounded by van hove singularities in the density of states corresponding to saddle points in the lowest energy minibands .
the weaker inversion - asymmetric terms , @xmath30 , in the second line of eq . ,
open a minigap in both types of mdp discussed above .
appearance of mdps at the edge of the first miniband results in a peculiar spectrum of electronic landau levels , as shown on the r.h.s of fig .
[ fig : fig2 ] .
each data point in these spectra represents one of the hofstadter minibands @xcite ( with an indistinguishably small width ) calculated for rational values of magnetic flux , @xmath77 per moir supercell following a method in ref .
@xcite . using these spectra
one can trace a clearly separated `` zero - energy '' landau level related to the isolated @xmath61-point mdp in the valence band in the bottom row of fig .
[ fig : fig2 ] , in addition to the true zero - energy landau level at the conduction - valence band edge .
the three isolated mdps on the sbz edge in the valence band ( top row of fig .
[ fig : fig2 ] ) also result in a `` zero - energy '' landau level , though not as clearly separated and split by the magnetic breakdown occurring at @xmath78 . to summarise , the inversion - symmetric moir perturbation will result in either the first sbz separated from the rest of the spectrum by one or three mdps , or , for weak perturbations , will result in overlapping first and higher minibands .
the experimental consequences of this can be expected in the optical spectroscopy of graphene on a hexagonal substrate : the presence of mdps and van hove singularities in the density of states should lead to a modulation of the fir and ir absorption spectra of monolayer graphene , which otherwise , has the flat absorption coefficient of 2.3% .
another experimental consequence of the moir minibands would consist in a non - monotonic variation of the hall coefficient upon doping the graphene flake with electrons or holes .
for example , for those miniband spectra in fig .
[ fig : fig2 ] , where there are isolated mdps in the valence band , the hall coefficient would pass through a zero value and change sign at two characteristic densities , @xmath79 and @xmath80 . at the density @xmath79 , which corresponds to the valence band filled with holes up to the van hove singularity ,
the hall coefficient will change sign from positive to negative . at the higher density , @xmath80 , which corresponds to a completely filled first miniband
, it would repeat the behaviour at the neutrality point changing sign from negative to positive .
such behaviour is expected to take place for the entire regions of the parametric space painted red , blue or green in fig .
[ fig : fig1 ] .
the relation between these two carrier densities for various types and strengths of moir perturbations is shown in fig .
[ fig : fig3 ] .
for the clear part of the parametric space for which we find substantial overlap between many moir minibands such alternations in the sign of the hall coefficient would be obscured by the competing contributions from the `` electron - like '' and `` hole - like '' branches in the spectrum .
the authors thank f. guinea , a. macdonald , e. j. mele and p. san - jose for useful discussions during the 2012 kitp programme _ the physics of graphene _ , where this study was started .
we acknowledge financial support from dtc nownano , erc advanced grant _ graphene and beyond _ , eu strep _ conceptgraphene _ , royal society wolfson research merit award and epsrc science and innovation award . |
there is a wealth of literature available on the highly isotropic nature of cosmic - rays ( crs ) observed at the earth ( see e.g. the references given in thoudam 2007 , hereafter paper i ) .
the cr anisotropy amplitude is only @xmath5 in the energy range of @xmath6 ( guillian et al .
2007 and references therein ) with the phase ( direction ) mainly found in the outer galaxy , particularly in the second quadrant of the galaxy .
the possible explanations for the anisotropy are generally beleived to be the global diffusion leakage of crs from the galaxy , the random nature of the cr sources in space - time and the effect of the local sources . in paper
i , the effect of the known local supernova remnants ( snrs ) has been studied in detail by giving more emphasis to the particle release time .
the study found that the observed anisotropy data favour the burst - like injection model if particles are released from the sources at an age of @xmath7 .
the continuous injection model gives an anisotropy which is too large to explain the observed data .
however , paper i considered the cr diffusion zone as an unbounded three - dimensional space which is actually too far from the real geometry of the galaxy .
the present work is a continuation of the earlier work , but considers the diffusion region as a flat cylindrical disc having both radial and the vertical boundaries . in the present study ,
the propagation of crs is assumed to follow the same diffusion equation given in paper i. the solution will be applied to local snrs and the results will be compared to those obtained in paper i for the burst - like model of particle injection .
in the diffusion model , neglecting convection , energy losses and particle losses due to nuclear interactions , the propagation of cr protons in the galaxy is given by the equation @xmath8 where @xmath9 is the differential number density , @xmath10 is the proton kinetic energy , @xmath11 with @xmath12 constant ( positive ) is the diffusion coefficient which is assumed to be spatially uniform in the galaxy and @xmath13 is the proton production rate .
the cr propagation region is assumed to be a cylindrical box bounded in both the radial and vertical directions , and our calculation takes into acount the exact location of the sources with respect to the earth .
inspite of the fact that the actual spatial distribution of observed snrs extent as far as @xmath14 from the galactic plane ( stupar et al .
2007 ) , most of the cr propagation studies assume the sources to be uniformly distributed in a thin disc of half - thickness @xmath15 .
such an approximation is valid in the study of global properties of galactic crs since majority of the sources are confined within @xmath16 from the plane .
but , in studies like the present one where the effects of nearby discrete sources are discussed , the actual position of the sources should be considered since , for example , for the same source distance @xmath17 we expect to see different cr fluxes at different source heights due to the presence of the vertical halo boundary .
our calculation will also assume that the sun is located on the galactic plane since our solar system is only @xmath18 away from the plane ( cohen 1995 ) .
the green s function @xmath19 of eq .
( 1 ) , i.e. the solution for a @xmath20-function source term @xmath21 can be found so that the general solution can be obtained as @xmath22 since the cr particles are assumed to be liberated at time @xmath23 , the equation for @xmath19 at @xmath24 becomes simply @xmath25 eq .
( 3 ) is solved using the proper boundary conditions and the continuity equations .
while solving , we consider the origin to be located at @xmath26 from the galactic center .
note that later on this point will represent the actual position of the source with respect to the observer .
then , the cr density at a point @xmath27 due to a point source [ which is positioned at @xmath26 from the galactic center ] with age @xmath28 , is obtained using eq .
( 2 ) as @xmath29\right\rbrace\nonumber\\ \times \displaystyle\sum_{k=1}^{\infty}\left\lbrace sin\left(\frac{k\pi(r - y_i)}{2r}\right)sin\left(\frac{k\pi(r - y_i-|y|)}{2r}\right)exp\left[-\frac{k^2\pi^2d(t - t_0)}{4r^2}\right]\right\rbrace\nonumber\\ \times \displaystyle\sum_{n=1}^{\infty}\left\lbrace sin\left(\frac{n\pi(h - z_i)}{2h}\right)sin\left(\frac{n\pi(h - z_i-|z|)}{2h}\right)exp\left[-\frac{n^2\pi^2d(t - t_0)}{4h^2}\right]\right\rbrace\end{aligned}\ ] ] where @xmath30 and @xmath0 represent the radial and the vertical boundaries of the galaxy respectively .
the solution at @xmath31 is obtained by just replacing @xmath26 with @xmath32 in eq .
the proton flux can be calculated using @xmath33 , where @xmath34 is the velocity of light and the source spectrum @xmath35 is taken as @xmath36 in which @xmath37 is the proton mass energy and @xmath38 is the normalization constant .
the source spectral index @xmath39 is chosen such that @xmath40 , the observed proton spectral index ( haino et al .
2004 ) . for very large radial boundary @xmath41 , the solution of eq .
( 1 ) at @xmath42 can be written as @xmath43
\displaystyle\sum_{n=1}^{\infty}\left\lbrace sin\left(\frac{n\pi(h - z_i)}{2h}\right)sin\left(\frac{n\pi(h - z_i-|z|)}{2h}\right)exp\left[-\frac{n^2\pi^2d(t - t_0)}{4h^2}\right]\right\rbrace\ ] ] fig .
1 compares the proton flux at the galactic center given by eq .
( 4 ) with that of eq . ( 6 ) for an snr - like source located at @xmath44 away from the center with an age @xmath45 .
the results of eq .
( 4 ) at @xmath46 are shown by the thin solid , dashed and dotted lines respectively .
the thick solid lines represent the unbounded solution given by eq .
( 6 ) ( i.e. the solution for @xmath47 )
. the calculations are done at @xmath48 and at @xmath49 assuming @xmath50 , represented by the left- and right - hand figures respectively .
the diffusion coefficient is taken as @xmath51 for @xmath52 , where @xmath10 is in gev ( engelmann et al .
1990 ) and the injected protons are assumed to carry @xmath53 percent of the total explosion energy of @xmath54 .
the figures clearly show that , for sources near to the observer , the solution of eq .
( 4 ) can be very well approximated by the much simpler unbounded solution for any value of @xmath0 if @xmath55 .
for example , the results at @xmath56 exactly coincide with the @xmath47 lines .
therefore , considering the fact that our solar system is positioned at a distance of @xmath57 from the galactic center and that the galactic radius extends as far as @xmath58 , the effect of the radial boundary @xmath30 on the observed crs should be negligible at least for those sources that can give appreciable density fluctuations at the earth , i.e. for those sources located within @xmath59 from the earth ( see thoudam 2006a ) . in the following sections where we study the effect of nearby snrs on the observed crs , we will therefore adopt the simpler eq .
( 6 ) instead of the complicated eq .
( 4 ) .
cr proton flux at the galactic center due to an snr - like source with age @xmath60 located at @xmath44 away from the center .
the thin solid line , dashed and dotted lines are the fluxes calculated using eq .
( 4 ) for @xmath46 respectively .
the thick solid line represents the flux calculated for the boundaryless case ( @xmath47 ) using eq .
the calculation assumes @xmath50 and @xmath61 .
_ left _ : for @xmath48 .
_ right _ : @xmath49 . from the figures
, it can be seen that for any value of @xmath0 eq .
( 4 ) can be well approximated by the much simpler boundaryless solution if @xmath55.,title="fig:",scaledwidth=28.0% ] cr proton flux at the galactic center due to an snr - like source with age @xmath60 located at @xmath44 away from the center .
the thin solid line , dashed and dotted lines are the fluxes calculated using eq .
( 4 ) for @xmath46 respectively .
the thick solid line represents the flux calculated for the boundaryless case ( @xmath47 ) using eq .
the calculation assumes @xmath50 and @xmath61 .
_ left _ : for @xmath48 .
_ right _ : @xmath49 . from the figures
, it can be seen that for any value of @xmath0 eq .
( 4 ) can be well approximated by the much simpler boundaryless solution if @xmath55.,title="fig:",scaledwidth=28.0% ]
knowing the cr density at a point @xmath62 away from a source of age @xmath28 , the single source anisotropy amplitude in the diffusion approximation can be calculated using ( mao @xmath63 shen 1972 ) @xmath64 where @xmath65 is given by eq .
( 6 ) for a point source @xmath66 located at @xmath26 from the earth .
the total anisotropy parameter at the earth due to a number of nearby discrete sources in the presence of an isotropic cr background is given by ( paper i ) @xmath67 where the summation is over the nearby discrete sources .
@xmath68 denotes the direction of the source @xmath66 giving a flux @xmath69 and @xmath70 denotes the direction of maximum intensity .
@xmath71 represents the total observed flux of cr protons above @xmath72 ( haino et al .
the phase of the anisotropy is taken as the direction of maximum intensity .
therefore , the anisotropy @xmath20 as well as the phase at an energy @xmath10 depends on the age and distance of the nearby sources , and may be determined by different sources at different energy intervals . however , in the case of a single source dominance , the total anisotropy @xmath20 is given by @xmath73 , where @xmath74 denotes the source giving the maximum flux at the earth .
cr anisotropy at the earth due to nearby known snrs assuming burst - like injection model . figs .
( a ) , ( b ) , ( c ) @xmath63 ( d ) are the results obtained for @xmath75 @xmath63 @xmath76 respectively .
thin solid lines represent the results of eq .
( 8) for @xmath77 , the thin dashed lines for @xmath48 , dotted lines for @xmath78 and the dot - dashed lines for @xmath49 .
the thick solid lines are the results of paper i i.e. for @xmath79
( a ) , ( b ) @xmath63 ( c ) show that different sources determine the anisotropy at different energy intervals .
these are marked by the source names along the lines . in fig .
( d ) , the monogem snr solely determine the anisotropy in the whole energy range .
the thick dashed line is the best fit result , in the case of a single source dominance , calculated assuming infinite boundaries .
data points are taken from the compilation of various results given in erlykin @xmath63 wolfendale 2006.,title="fig:",scaledwidth=28.0% ] cr anisotropy at the earth due to nearby known snrs assuming burst - like injection model . figs .
( a ) , ( b ) , ( c ) @xmath63 ( d ) are the results obtained for @xmath75 @xmath63 @xmath76 respectively .
thin solid lines represent the results of eq .
( 8) for @xmath77 , the thin dashed lines for @xmath48 , dotted lines for @xmath78 and the dot - dashed lines for @xmath49 .
the thick solid lines are the results of paper i i.e. for @xmath79
( a ) , ( b ) @xmath63 ( c ) show that different sources determine the anisotropy at different energy intervals .
these are marked by the source names along the lines . in fig .
( d ) , the monogem snr solely determine the anisotropy in the whole energy range .
the thick dashed line is the best fit result , in the case of a single source dominance , calculated assuming infinite boundaries .
data points are taken from the compilation of various results given in erlykin @xmath63 wolfendale 2006.,title="fig:",scaledwidth=28.0% ] cr anisotropy at the earth due to nearby known snrs assuming burst - like injection model
( a ) , ( b ) , ( c ) @xmath63 ( d ) are the results obtained for @xmath75 @xmath63 @xmath76 respectively .
thin solid lines represent the results of eq .
( 8) for @xmath77 , the thin dashed lines for @xmath48 , dotted lines for @xmath78 and the dot - dashed lines for @xmath49 .
the thick solid lines are the results of paper i i.e. for @xmath79 .
( a ) , ( b ) @xmath63 ( c ) show that different sources determine the anisotropy at different energy intervals .
these are marked by the source names along the lines . in fig .
( d ) , the monogem snr solely determine the anisotropy in the whole energy range .
the thick dashed line is the best fit result , in the case of a single source dominance , calculated assuming infinite boundaries .
data points are taken from the compilation of various results given in erlykin @xmath63 wolfendale 2006.,title="fig:",scaledwidth=28.0% ] cr anisotropy at the earth due to nearby known snrs assuming burst - like injection model .
( a ) , ( b ) , ( c ) @xmath63 ( d ) are the results obtained for @xmath75 @xmath63 @xmath76 respectively .
thin solid lines represent the results of eq .
( 8) for @xmath77 , the thin dashed lines for @xmath48 , dotted lines for @xmath78 and the dot - dashed lines for @xmath49 .
the thick solid lines are the results of paper i i.e. for @xmath79
( a ) , ( b ) @xmath63 ( c ) show that different sources determine the anisotropy at different energy intervals .
these are marked by the source names along the lines . in fig .
( d ) , the monogem snr solely determine the anisotropy in the whole energy range .
the thick dashed line is the best fit result , in the case of a single source dominance , calculated assuming infinite boundaries .
data points are taken from the compilation of various results given in erlykin @xmath63 wolfendale 2006.,title="fig:",scaledwidth=28.0% ] in this section , we will try to investigate whether the presence of a halo boundary can affect the anisotropy at the earth due to nearby sources . for that
, we consider the @xmath80 known snrs located within @xmath81 from the earth as listed in table 1 of paper i. the total anisotropy due to these snrs is calculated using eq .
( 8) for different @xmath0 values at different @xmath82 s .
2 shows the comparison of the anisotropies calculated in the present work with those obtained in paper i for the burst - like particle injection model .
the data points are taken from the compilation of various experiments given in erlykin @xmath63 wolfendale ( ew ) 2006 .
figs 2(a ) , ( b ) , ( c ) and ( d ) are the results obtained for @xmath75 @xmath63 @xmath76 , respectively .
the thin solid lines represent the results of eq .
( 8) for @xmath77 , the dashed lines are for @xmath48 , the dotted lines are for @xmath78 and the dot - dashed lines are for @xmath49 .
the thick solid lines are the results of paper i which were obtained assuming @xmath79 [ eq .
11 of paper i ] . in figs
2(a)@xmath4(c ) , different sources determine the anisotropy at different energy ranges .
these are marked by the source names along the lines .
it can be seen that the results for @xmath77 show a noticeable deviation from the @xmath79 lines , while those for @xmath48 show a very slight deviation .
the results for other higher @xmath0- values almost overlap with the @xmath79 lines and are not easily visible in the figures .
this shows that , for the particle release time of @xmath83 , the halo height effect on the local snr contribution to the observed cr anisotropy is almost negligible if @xmath84 .
however , the situation is somewhat different in fig .
2(d ) where the calculations are performed at @xmath85 . note that this value of particle injection time is that at which the model calculated anisotropy values are close to the observed data ( see the results of paper i ) .
the anisotropy here is determined solely by the monogem snr in the whole energy range considered here , and only those results for @xmath86 show considerable variation from the @xmath79 line .
the results for @xmath87 show a negligible deviation . combining all the results of fig .
2 , we can finally conclude that the effect of the halo boundary of our galaxy on the local snr contribution to the observed cr anisotropy is negligible as long as the boundary is greater than @xmath88 . in the next section
, we will combine this result along with the halo heights obtained by several authors to discuss the importance of @xmath0 in the anisotropy study due to local sources .
the effect of the nearby cr sources is considered as one of the important effects that can give rise to the observed cr anisotropy at the earth .
however , the calculation of cr fluxes from any type of source in the galaxy essentially requires the use of the proper geometry of the galaxy as well as the actual position of the source with respect to the observer .
since our galaxy has a cylindrical geometry with the radius much larger than the height , the radial boundary is found to have a negligible effect on the cr density and hence the geometry can be approximated by an infinite radius with a finite vertical height .
furthermore , this study has found that the effect of the vertical halo boundary on the local snr contribution to the cr anisotropy is negligible if @xmath89 .
2 shows the effect of the halo height on the cr anisotropy due to nearby known sources for different particle injection times . among the 13 snrs considered , only monogem , vela , g299.2 - 2.9 , sn185 and cygnus loop
are found to determine the anisotropy at different energy intervals . also , all of them except sn185 ( with @xmath90 ) have distances @xmath91 .
this shows that only the nearest sources mainly determine the anisotropy as expected , and hence this results in a negligible halo height effect for @xmath87 .
it is also worth mentioning that the vertical heights of the dominant sources above the galactic plane are found to be less than @xmath92 which is much less than the halo heights @xmath93 considered here .
the actual value of the halo height of our galaxy is not exactly known .
its value is generally obtained along with other propagation parameters using the observed cr data like the secondary@xmath94primary ratios , cr density distribution etc .
but , the values obtained from the same experimental data are different for different cr propagation models .
webber , lee @xmath63 gupta ( 1992 ) had obtained a value of @xmath95 using diffusion - convection model .
lukasiak et al .
( 1994 ) had obtained @xmath96 using the webber et al .
( 1992 ) model without convection .
webber @xmath63 soutoul ( 1998 ) obtained @xmath97 and @xmath98 using the diffusion and monte carlo models respectively .
other results like those of freedman et al .
( 1980 ) and ptuskin @xmath63 soutoul ( 1998 ) obtained @xmath99 and @xmath100 respectively . a completely numerical approach using more realistic physical conditions of the galaxy determined a value of @xmath101 for the diffusion - convection model and @xmath102 for the re - acceleration model ( strong @xmath63 moskalenko 1998 ) .
these results are found to be consistent with the observations of galactic radio emission structure at 408 mhz which indicate the presence of a thick radio disk with full equivalent width of @xmath103 , @xmath104 and @xmath105 in the galactic radial range of @xmath106 , @xmath107 and @xmath108 respectively ( beuermann et al .
1985 ) , but such a wide range of values makes the galactic halo height a very uncertain parameter in cr propagation studies .
however , since most of the values obtained are found to have @xmath1 , the conclusion given in the previous section suggests that the study of local crs due to nearby snrs can be carried out without having much information on @xmath0 .
this is because the effect of the nearest sources @xmath109 dominates over the influence of the other nearby sources and the cr fluxes from these sources are almost independent of the halo boundary for @xmath89 as discussed before .
hence , the study of the effect of local sources on the cr anisotropy at the earth can be done using the much simpler three - dimensional unbounded solution .
for the infinite boundary case , if a single source dominates the anisotropy in the whole energy range as in fig .
2(d ) , the total anisotropy follows an energy dependence of the form @xmath110 in the high energy regime ( paper i ) , which for @xmath111 goes as @xmath112 .
such a decrease with energy is in fact observed in the high energy anisotropy data somewhere above @xmath113 upto around @xmath114 .
moreover , the increase in anisotropy from @xmath5 in the @xmath115 energy range can also be possibly explained by a proper choice of @xmath116 or rather @xmath117 for the single dominant source .
we try to estimate the physical parameters of such a source that best fit the data .
the best - fitting parameters are found to be @xmath118 and @xmath119 , and the best - fitting line is shown as the thick dashed line in fig .
thus , for @xmath50 the source should have an age of @xmath120 . however , it is possible to obtain a number of @xmath121 combinations which equally fit the data , all of them giving the same value of @xmath122 and @xmath123 .
therefore , the present study only gives an estimate of the distance to the single dominant source ; it does not give any precise information on the age and the particle release time of the source .
it should be noted that it is not the individual @xmath121 values that determine the contribution of the source , but the propagation time @xmath123 of the particles after their release from the source .
we can determine the best - fitting @xmath28- value only if we know @xmath82 , but the value of @xmath82 is not exactly known .
it may even be that @xmath82 is an energy - dependent parameter , i.e. particles with different energies emitted at different times .
studies based on diffusive shock acceleration in snrs have shown that the highest energy particles start leaving the source region already at the beginning of the sedov phase ( berezhko et al .
1996 ) , but the major fraction of accelerated crs remain confined for almost around @xmath124 for an interstellar medium ( ism ) hydrogen atom density of @xmath125 .
this implies that for the local ism which has @xmath126 ( see e.g. thoudam 2006b and references therein ) , if a single source determines the whole anisotropy , the source should have a characteristic age of @xmath2 .
unfortunately , there is no nearby known snr with such an age located at @xmath127 .
however , it is quite possible that the single dominant source may be an _ undetected _ old snr .
in fact , studies assuming adiabatic phase in snr evolution have shown that the surface brightness of an snr of age @xmath128 yrs lies below the detection limit of radio telescopes ( leahy @xmath63 xinji 1989 ) .
the present result is further supported by the fact that almost all the nearby sources are quite young with estimated ages less than @xmath129 ( the generally accepted particle release time ) , and they might not have released the crs into the local ism . in addition , the possiblity that some of the observed features of crs may be due to _ undetected _ nearby sources can not be simply ignored .
the single - source explanation of the observed cr properties can also be found in some earlier works ( e.g. ew 2000 and references therein ; ew 2006 , etc . ) , but in a somewhat different context .
ew 2000 claimed that the knee in the cr spectrum at @xmath130 can be attributed to the presence of a single recent supernova ( as yet unidentified ) in the local region . on the other hand ,
ew 2006 tried to explain the rise in the anisotropy amplitude as well as the change in its phase near the knee using a single source exploded in the direction from the sun downward of the main cr flux , which are predominantly coming from the inner galaxy .
the latter study considered the source parameters as similar to those of the monogem snr .
although the single source idea has not been readily accepted by the cr community , at the same time there is no reason why it should be just neglected .
the present study even points out one more observed property of crs that can possibly be explained by the single source model .
99 berezhko , e. g. , yelshin v.k .
@xmath63 ksenofontov l.t .
1996 , j. exp .
phys . , 82 , 1 beuermann , k. , kanbach , g. , @xmath63 berkhuijsen , e. m. 1985 , a@xmath63a , 153 , 17 cohen , m. , 1995 , apj , 444 , 874 engelmann , j. j. , ferrando , p. , soutoul , a. , goret , p. , @xmath63 juliusson , e. 1990 , a@xmath63a , 233 , 96 erlykin , a. d. , @xmath63 wolfendale , a. w. 2000 , a@xmath63a , 356 , l63 erlykin , a. d. , @xmath63 wolfendale , a. w. 2006 , astropart .
phys . , 25 , 183 freedman , i. , kearsey , s. , osborne , j. l. , @xmath63 giler , m. 1980 , a@xmath63a , 82 , 110 guillian , g. , et al .
2007 , phys .
d , 75 , 062003 haino , s. , et al .
2004 , phys .
b594 , 35 leahy , d. a. , @xmath63 xinji , w. 1989 , pasp , 101 , 607 lukasiak , a. , ferrando , p. , mcdonald f. b. , @xmath63 webber , w. r. 1994 , apj , 423 , 426 mao , c. y. , @xmath63 shen , c. s. 1972 , chinese j. phys .
, 10 , 16 ptuskin , v. s. , @xmath63 soutoul , a. 1998 , a@xmath63a , 337 , 859 strong , a. w. , @xmath63 moskalenko , i. v. 1998 , apj , 509 212 stupar , m. , filipovi@xmath131 , m. d. , parker , q. a. , white , g. l. , pannuti , t. g. , @xmath63 jones , p. a. 2007 , ap@xmath63ss , 307 , 423 thoudam , s. 2006a , mnras , 370 , 263 thoudam , s. 2006b , astropart .
phys . , 25 , 328 thoudam , s. 2007 , mnras , 378 , 48 ( paper i ) webber , w. r. , lee , m. a. , @xmath63 gupta , m. 1992 , apj , 390 , 96 webber , w. r. , @xmath63 soutoul , a. 1998 , apj , 506 , 335 | in an earlier paper , the effect of the nearby known supernova remnants ( snrs ) on the local cosmic - rays ( crs ) was studied , considering different possible forms of the particle injection time .
the present work is a continuation of the previous work , but assumes a more realistic model of cr propagation in the galaxy .
the previous work assumed an unbounded three - dimensional diffusion region , whereas the present one considers a flat cylindrical disc bounded in both the radial and vertical directions .
the study has found that the effect of the vertical halo boundary @xmath0 on the local snr contribution to the observed cr anisotropy is negligible as long as @xmath1 .
considering the values of the halo height @xmath1 obtained by different authors , the present work suggests that the study of the effect of local sources on the cr anisotropy can be carried out without having much information on @xmath0 and hence , using the much simpler three - dimentional unbounded solution .
finally , the present work discusses about the possibility of explaining the observed anisotropy below the knee by a single dominant source with properly chosen source parameters , and claims that the source may be an _ undetected _ old snr with a characteristic age of @xmath2 located at a distance of @xmath3 from the sun .
[ firstpage ] cosmic rays@xmath4supernova remnants |
works exploring the quantum effects on multi - photon spatial interference and images have traditionally focused on experiments using a pair of photons @xcite .
other important results obtained in the past decades , involving the spatial degree of freedom of photon pairs , were the complementarity between entanglement and single - photon coherence @xcite , increased resolution on imaging when using entangled photons @xcite , the spatial quantum eraser @xcite , measurement of the two - photon de broglie wavelength @xcite , and the experiments encoding logical qubits and qudits onto the path state of photons going through slits @xcite . in this paper
we demonstrate control over the quantum correlations of two biphotons that are encoded in a four qubit space by using a double slit .
our setup is able to generate two pairs of highly entangled photons in one configuration and four nearly - separable photons in another .
we further present a way of characterizing the state generated , without having to resort to fourfold coincidence detection , by exploiting the fact that each pair contains one photon of horizontal polarization and one of vertical polarization .
this allows us to observe both the contribution that is common to single and double pair spontaneous parametric down - conversion ( spdc ) events and the one that is exclusive to double pair events , i.e. coincidences between photons with the same polarization .
finally , we also observed that the diffraction behavior depended on the quantum correlations of the photons , producing an asymmetrical pattern in the case of entangled photons and the expected diffraction pattern for the case of photons in a nearly - separable spatial state . by encoding the transverse profile of photons on a discrete space with the use of slits
, it becomes simple to scale the dimensionality , from qubits to qudits , by using more slits .
this experiment , then , opens a path for future works using multi - pair qudit states .
as it is customary , double - slits and multi - slits are characterized by their width @xmath1 and by the center - to - center distance @xmath2 .
in order to exploit the possibility of scaling experiments to include quantum systems of dimensions greater than 2 , it has become usual to encode logical qubits and qudits on the transverse momentum and/or position state of photons .
our experiment encodes the logical qubits @xmath3 and @xmath4 on the condition that the photon passes through the left or right slit , respectively .
as presented for the two - photon case in @xcite , the most general state of a pair of qubits generated using a double slit and assuming a symmetrical illumination can be written as : @xmath5 where @xmath6 and @xmath7 are maximally entangled states of two photons a and b known as bell states : @xmath8 up to now , experiments have been restricted to controlling the spatial state of a single bi - photon , by using linear optical elements . in this work
we demonstrate experimental control of quantum correlations in a four - photon state comprised of two pairs of photons . by using a pulsed pump and a high - efficiency periodically - poled ktiop@xmath9 crystal @xcite , the second order term in @xmath10 on the expansion of the time - evolution operator associated to
the spdc becomes relevant . in our experimental conditions
, this term results in the generation of two distinguishable pairs of photons .
the pump is a pulsed beam obtained by second harmonic generation on a bbo crystal pumped by an femtosecond pulsed laser that has pulses of 200 fs and a repetition rate of .
the mean power of the pump beam when falling on the crystal is of the order of .
it is loosely focused by a spherical lens onto a long , type - ii ppktp grown for collinear and degenerate spdc centered on @xmath11 .
the created pair will consist of one horizontally ( h ) and one vertically ( v ) polarized photon .
polarization was used as an auxiliary mode , in order to separate the photons of each pair into two distinct physical subsystems a and b , much in the same way non - collinear experiments that encode the qubits using polarization use the down - conversion linear momentum modes .
the general state , considering one or two pairs cross the double - slit is , up to appropriate normalization @xcite , @xmath12 where @xmath13 is the generation efficiency , depending on both the non - linear susceptibility and the pump intensity , and @xmath14 is a normalization constant .
the four - photon component can then be written as @xmath15 since there is no genuine multipartite entanglement , it is sufficient to characterize the entanglement by the concurrence @xcite of the single pair , which is given by @xcite @xmath16 states with @xmath17 or @xmath18 are maximally entangled .
conversely , states with @xmath19 and @xmath20 are separable .
the presence of two photons of horizontal and vertical polarization also gives us the opportunity to consider the partial state of same - polarization photons .
in practical terms , this is obtained by using a polarized beamsplitter to create two detection branches , one for horizontally and one for vertically polarized photons .
since the two pairs are not entangled , the purity of the partial state , defined as @xmath21 and given below , will be used as a relevant indicator on those same - polarization coincidences , as it is directly related to the degree of entanglement in each pair : @xmath22 the combination of eq .
( [ concurrence ] ) and eq .
( [ purity ] ) allows us to write @xmath23
in order to control the quantum correlations in the four photons , we will resort to a tunable linear optical scheme . here
we present two extremal conditions , that is , the pairs of photons coming out of the double slit are either highly entangled or separable . the correlations after the double - slit depend on how the bi - photon field generated inside the crystal is projected onto the plane of the apertures .
influence over the state encoded on the slits can be exerted either by controlling the pump field , the propagation of the bi - photons from the crystal to the slits or both .
since the bi - photon field is not affected in one of the transverse dimensions by the double - slit ( the larger dimension , i.e. the @xmath24 direction ) , we restrict our analysis to the @xmath25 direction ( the smallest slit dimension ) .
more specifically , the two - photon field in the crystal can be written , in the transverse momentum variables , as @xcite @xmath26where @xmath27 are the transverse components of the momentum of photon 1 and 2 in the bi - photon .
the tilde indicates that the functions refer to the transverse momentum space .
therefore @xmath28 is the angular spectrum of the pump beam and @xmath29 is the phase - matching condition for the ppktp . in most cases , including our experiment , @xmath28 is the angular spectrum of a gaussian beam .
hence , @xmath30 when the beam waist , @xmath31 , is in the center of the crystal , with @xmath32 corresponding to the fourier transform and @xmath33 and @xmath34 to the transverse position of photons 1 and 2 .
if the last condition is not satisfied , phase components due to the curvature of the wavefront must be considered as well . for our experimental conditions , the phase matching function is @xmath35 where @xmath36 is a phase - mismatch term that we consider to be negligible , @xmath37 is the frequency of the degenerate down - converted photons and the effective refractive index @xmath38 takes into account the fact that the two photons in type - ii spdc propagate differently inside a crystal of length l. a more thorough description of the phase - matching function can be found in @xcite .
+ figure [ setup ] shows the experimental setups used in this work to control the quantum correlations of the bi - photons at the double - slit plane .
we adopted two configurations : to generate highly entangled states and to prepare separable states .
scheme projects a magnified image of the crystal plane onto the double - slit .
this geometry is achieved by adopting two lenses , a cylinder lens with focal lens @xmath39 and a spherical one with @xmath40 , so the magnification factor of the image reads @xmath41 .
we adopt a cylinder lens in order to create a crystal image in the direction perpendicular to the slits , the x - direction , while keeping the far - field image in the y orthogonal direction . by such approach the dominant contribution to the bi - photon amplitude transmitted by the double - slit
is the one associated to the phase - matching function .
that is so because , in the magnified image of the crystal center projected on the slits , the gaussian profile of the pump component is much wider than the fourier transform of the @xmath42 function of the phase - matching condition and can be considered constant .
configuration relies on the far - field imaging onto the double slit .
this scheme is obtained by adopting a single spherical lens with focal length @xmath43 placed at a distance @xmath44 from both the center of the crystal and the plane of the slits . with this geometry
the bi - photon amplitude is almost constant along the double - slit apertures leading to the generation of separable states . a detailed discussion of the relation between the parameters @xmath45 and @xmath46 of eq .
( [ general ] ) and the adopted geometries can be found in @xcite , in the form of what was called an engineering parameter @xmath47:@xmath48 that gives a clear method for estimating the values of @xmath45 and @xmath46 by knowing the form of the two - photon amplitude function @xmath49 specifically when @xmath50 and @xmath51 , where @xmath33 and @xmath34 are coordinates on the plane of the double - slit and @xmath52 , as previously mentioned , is the distance from the center of one of the slits to the center of the slit pattern . in this case , the distance between the center of the slits is @xmath2 .
we refer to the values calculated through that method when giving expected values for @xmath45 and @xmath46 . for our experimental setups ,
the expected values are @xmath53 and @xmath54 in the configuration of fig . .
by changing to the configuration of fig .
we expect @xmath55 and @xmath56 .
let us now discuss the measurements necessary to characterize the states generated by the setups presented in fig .
[ setup ] .
the parameter @xmath45 can be inferred by measurements on the image of the slits , which corresponds to the @xmath57 basis .
measurements for determining @xmath46 , however , require us to go to the diagonal basis . for
qubits encoded on transverse spatial variables , this can be achieved by looking at coincidence measurements on the pattern generated in the far field of the slits , that gives information on both @xmath45 and @xmath46 .
one way to access the far field is to use a lens with focal length @xmath58 in the @xmath44-@xmath44 configuration .
this maps the fourier transform of the slits onto positions on the detection plane , as shown in eq .
5 - 19 of @xcite .
the probability of coincident detection of two photons from the same pair , one of horizontal polarization and the other of vertical polarization , when using the far - field strategy , can be derived as a function of the position of both detectors involved , @xmath59 . to calculate it
, we first assume the contribution of only a single pair of photons and then follow mandel and wolf @xcite to obtain @xmath60 where @xmath61 , for @xmath62 and @xmath63 is calculated by propagating the electrical field operators from the plane of the slits to the detection plane , taking into account the lens placed at a distance @xmath58 from both planes .
this , as mentioned before , means taking the optical fourier transform of the product of the aperture function and the amplitude of the bi - photon on the plane of the slits @xcite . after performing the calculations
, we obtain the explicit form of @xmath64 : @xmath65 + \nonumber \\ & \sin ^2(\alpha/2)\cos(\beta(x_1+x_2)).\label{pdbc}\end{aligned}\ ] ] this function can be represented on a plane , which we will call the coincidence map .
the same approach can be adopted to represent the coincidence counts detected in the experiment , allowing us to compare experimental and theoretical maps .
we see that it depends of both @xmath45 and @xmath46 , meaning the coincidence maps can be used to characterize both parameters of the state .
the following step is to characterize the measurements sensitive exclusively to the two - pair contribution @xmath66 .
since we work with two pairs of photons , the most natural type of measurement would be fourfold coincidence detection .
this , however , is impractical due to the fact that only a small fraction of the total signal is collected by the detectors when in the far - field regime .
as fourfold coincidences depend on the overall efficiency of four detectors , it can require extremely long detection times .
to circumvent this predicament , we use the fact that our source of photon pairs is a type - ii crystal and , thus , coincidences on the same polarization branch can only come from down - conversion events in which two pairs were created .
another factor that favors same - branch coincidences as a method for investigating the two - pair signature is that only one beamsplitter ( bs ) is involved , whereas two are necessary on fourfold detection events , as can be seen in fig .
[ detection ] . these coincidences can be mapped in much the same way as those from two photons with orthogonal polarization .
even though double - pair events are not the major contribution , the probability of detecting h - h or v - v coincidences is still much larger than of detecting fourfold coincidences , since only two detectors are involved in the process and , therefore , only two detection efficiencies are taken into account .
the expression for coincidence detection of photons on the same polarization branch is : @xmath67 to calculate it , we use @xmath68 in which @xmath69 can be either @xmath70 or @xmath71 .
indexes @xmath72 and @xmath73 correspond to the two pairs of photons .
the density matrix @xmath74 is given by @xmath75 , with @xmath76 being the partial trace over the photon of polarization @xmath77 .
note that the same derivation can be used to calculate the probability of detecting coincidences between photons of different polarization and different pairs .
equations ( [ pdbc ] ) and ( [ psbc ] ) have been derived under the assumption that the incident two - photon field can be considered constant over the width of a single slit , which is taken to be very narrow .
this allowed us to factor out the diffraction term as an enveloping function that is radially symmetric around the origin in the plane formed by @xmath33 and @xmath34 . as we will discuss in section [ diffraction ] , this is not true for all situations , but is a valid approximation if one is interested only in studying , for instance , the interference behavior close to the center of the detection plane .
we will now discuss in detail the experimental procedure used to obtain the coincidence maps .
the detection setup , shown in fig .
[ detection ] is composed of a spherical lens fl of focus length @xmath78 and four avalanche photo - diodes ( apd ) detectors divided into two polarization branches , as indicated above by the dashed bounding boxes .
each branch corresponds to one of the outputs of a polarized beamsplitter ( pbs ) and contains a balanced beamsplitter ( bs1 and bs2 ) .
thus , coincidences can be detected between apds on the same polarization branch , to which we refer as * sbc * ( * * s**ame * * b**ranch * * c**oincidences ) , and on different branches , referred to as * dbc * ( * * d**ifferent * * b**ranch * * c**oincidences ) .
each detector is mounted on a manual translation stage in the x direction , defined as being parallel to the optical table plane and orthogonal to the propagation direction of the beam .
additionally , a pinhole and an interference filter ( @xmath79 ) centered on are placed in front of the apd .
coincidences are detected in a 1 ns window using a custom electronic circuit .
such a time frame is large enough to collect all events originated from a single pulse but small enough so that there can be no coincidence between photons from two different pulses , allowing us to consider each pulse as an isolated spdc event that may generate either zero , one or two pairs of photons .
the maps are the result of 21 scans ( 17 in the case of fig .
[ entangled_diffraction ] ) where two detectors , one in each branch remain in a fixed position and the other two are moved in steps of ( for fine scans over the main interference region ) or for broad scans , meant for observing the diffraction pattern over a larger area .
in addition to generating the maps , the data collected was also used to evaluate efficiency of our source . to calculate the contribution of single and double pair events , we have estimated the detection efficiency of our apparatus starting from the measured count and coincidence rates of detectors .
considering uncorrelated spatial photon pairs generated with an average pair number per pulse @xmath47 , detected with an overall efficiency @xmath80 , for an experiment with time duration @xmath81 and repetition rate @xmath82 , the average counts of the detector are equal to @xmath83 , while the measured coincidences are @xmath84 .
the value of @xmath80 can then be obtained as @xmath85 .
spatial correlations between the two photons can modify the previous relation , so the data used corresponds to the maxima of detection on each map .
following the experimental procedures detailed in the previous section , we obtain coincidence maps on the far field of the double slits , which will serve as fingerprints for identifying the states generated .
the parameters obtained from these maps were then compared to those predicted by the theoretical model of our experiment , as presented in section [ control ] .
maps on fig .
[ correlated ] , [ uncorrelated ] and [ diffraction_separable ] were generated from a matrix of 21x21 experimental points .
those on fig .
[ entangled_diffraction ] originate from a 17x17 grid , although the separation between points are the same as that of fig .
[ diffraction_separable ] , meaning that a smaller field of view was imaged , from -4 to instead of -5 to .
data points have been normalized relative to the maximal values of each map , for ease of comparison to the simulated maps based on the theoretical model from eq .
( [ pdbc ] ) and eq .
( [ psbc ] ) .
additionally , the latter are generated over the same grids of points , corresponding to the same values of @xmath33 and @xmath34 .
this means that artifacts due to interpolation are not exclusive to the experimental maps and should present themselves in the same manner on the simulations , minimizing discrepancies that may influence the interpretation of the processed data .
there are , however , some artifacts present only on experimental maps , due to noise on the data set . on the case of maps on fig .
[ correlated ] and fig .
[ uncorrelated ] , the dominant aspect is noise , since the structures are much larger than the spacing on the grid . on fig .
[ entangled_diffraction ] and fig .
[ diffraction_separable ] , however , the size of the finer structures is comparable to that of the spacing between points on the grid .
this makes the aliasing contribution more significant .
however , in those two figures the most interesting structures , those due to diffraction , are larger than the grid spacing .
equation ( [ pdbc ] ) was used to simulate the different branch coincidence map for comparing with the experimental dbc map .
since this equation does nt take in account coincidences between photons of orthogonal polarization but from different pairs , we have subtracted that contribution from @xmath86 by estimating the accidental coincidences as @xmath87 .
these accidental coincidences , calculated for each measured point , were also subtracted from the maps . using the corrected data
, we also characterize our source . calculating the average pair number @xmath88 we obtain @xmath89 and @xmath90 , respectively for the setup of fig . and .
we attribute the difference between the two values to the different geometries and to the spatial correlation observed with setup .
the effective single pass parametric gain , calculated after the double slit , is 0.65 for setup and 0.87 for .
the probabilities @xmath91 of generating @xmath92 pairs are , for and , @xmath93 , @xmath94 , @xmath95 , @xmath96 , @xmath97 and @xmath98 .
we now examine in detail the results of the entangled state obtained by using the setup of fig . and , after it , the near - separable state generated by the setup indicated in . in fig .
[ correlated ] , the sbc map , originated only by multi - pair events , confirms that the partial state is indeed mixed .
the fact that a small degree of structure , still present on the simulated sbc map , is not observed experimentally is attributed to detection fluctuations that mask the low visibility pattern .
the interference pattern observed on experimental dbc maps , which reflect mainly the single - pair contribution , is consistent with near maximal entanglement on the qubit picture .
those two results are interpreted as a sign that two distinct pairs of photonic spatial qubits with high purity and path entanglement have been generated . ) .
the lower row , from * ( d ) * to * ( f ) * , corresponds respectively to the coincidences between the h and v branches ( d2d3 ) and to the same map corrected by subtracting the uncorrelated contribution ( d2d3-d2*d3 ) , as well as a simulated dbc map , generated with eq .
( [ pdbc ] ) .
the simulation parameters used are @xmath99 and @xmath100 , corresponding to @xmath101 . for a measurement time of per point
, the maximum in map * ( a ) * corresponds to 549 coincidence counts and in map * ( d ) * to 1630 coincidence counts .
the measured maps were obtained at the fourier plane of fig .
[ detection ] , after the photon pairs cross the double - slit . ] in fig .
[ correlated](d ) and fig .
[ correlated](e ) one can notice that the visibility increases substantially after subtracting from the total detections the coincidences between photons of different pairs generated by the same pulse , and transmitted through the double - slit .
our result then demonstrates that highly entangled pairs can be generated even in the regime where many bi - photons are produced by a single pump pulse .
we also note that , on each of the three fringes visible in the pattern , the maxima are situated on @xmath102 or .
this period of coincides with the step used on the measurements of fig .
[ entangled_diffraction ] .
as previously stated , the expected values for the two characteristic parameters for the state were @xmath53 and @xmath54 . in order to maximize the agreement between the experimental data and the simulation , those two parameters
were changed to @xmath99 and @xmath100 .
such a change for @xmath46 would be expected if the position of the pump waist and of the focal plane of the cyllindrical lens differed by approximately 200 microns , which is well within the experimental uncertainties of our setup . for the values used in the simulation ,
the concurrence is @xmath101 and the purity for the partial state is @xmath103 , compared to a minimum of 0.25 for a maximally mixed state .
let us now analyze the case of the two - photon coincidence maps generated by the setup of fig .
, for which the theoretical prediction is a four - photon state of negligible entanglement in the path variables , after the double - slit . in this case
, the single- and double - pair contributions can be considered as a tensor product of two and four distinguishable single - photon pure path states , respectively .
this is evidenced by the fact that the experimental maps for sbc and dbc are similar . for a state of two pure and separable pairs ,
the partial state detected will also be pure .
conversely , for maximally entangled path states the partial state is maximally mixed .
thus , we see no interference on the map of fig .
[ correlated](a ) , but the maps of fig .
[ uncorrelated](a ) and fig .
[ uncorrelated](d ) are equal .
this is in direct relation to what was shown in @xcite . for this state ,
the expected values of the two characteristic parameters were @xmath55 and @xmath56 .
as done during our analysis of the entangled case , those two parameters were changed to @xmath104 and @xmath105 in the simulated maps .
such a change for both @xmath45 and @xmath46 is , again , well within the experimental uncertainties of our setup .
for the values used in the simulation , the concurrence is @xmath106 , which results in a high purity for the partial state , @xmath107 . ) .
the lower row , from * ( d ) * to * ( f ) * , corresponds respectively to the coincidences between the h and v branches ( d2d3 ) and to the same map corrected by subtracting the uncorrelated contribution ( d2d3-d2*d3 ) , as well as a simulated dbc map , generated with eq .
( [ pdbc ] ) .
the simulation parameters used are @xmath104 and @xmath105 , corresponding to @xmath106 . for a measurement time of per point
, the maximum in map * ( a ) * corresponds to 122 counts and in map * ( d ) * to 247 counts .
the measured maps were obtained at the fourier plane of fig .
[ detection ] , after the photon pairs cross the double - slit . ]
one should note that the fluctuations observed in fig .
[ uncorrelated ] are due to the lower statistics , compared to those of fig .
[ correlated ] . in the former ,
poissonian fluctuations are more significant . another important point to be observed about the maps in fig .
[ uncorrelated ] is that the peaks are away from each other , twice the step used on fig .
[ diffraction_separable ] .
in order to fully understand our experiment , we now turn our attention briefly to the subject of the diffraction envelope .
we present a qualitative discussion of its characteristics for the two extremal cases of transverse spatial quantum correlations that were generated in this work . as in the previous section
, we find it appropriate to discuss the case of a highly entangled state and of a near - separable state separately .
figure [ entangled_diffraction ] shows a larger area for the same experimental conditions as fig . [
correlated ] .
the asymmetric nature of the diffraction envelope for different branch coincidences is clear , with a larger diffraction on the @xmath108 direction and a narrower one for @xmath109 , if compared to the expected homogeneous envelope from previous models concerning diffraction for the same type of experiment @xcite .
moreover , the diffraction on the case of same branch coincidences is even larger than predicted by models that consider the amplitude to be constant over a slit . ) and eq .
( [ sincpm ] ) .
the lower row , from * ( d ) * to * ( f ) * , corresponds respectively to the coincidences between the h and v branches ( d2d3 ) and to the same map corrected by subtracting the uncorrelated contribution ( d2d3-d2*d3 ) , as well as a simulated dbc map , using eq .
( [ pdbc ] ) and eq .
( [ sincpm ] ) .
the simulation parameters used are , as before , @xmath99 and @xmath100 . for a measurement time of per point
, the maximum in map * ( a ) * corresponds to 483 counts and in map * ( d ) * to 1598 counts . ] with the purpose of understanding this , we simulated the diffraction through a single slit for the setup of fig .
, by calculating numerically the optical fourier transform of the product of the slit transmission function and the amplitude of the bi - photon .
this was done on the @xmath110 and @xmath111 directions , which are those that exhibit interesting properties for the different branch coincidences .
we show below the optical fourier transforms used , based once more on eq . 5 - 19 of @xcite : @xmath112dx_1dx_2,\label{sincpm}\ ] ] where @xmath113 is the wave number of the down - converted photons , @xmath114 are coordinates on the plane of the slits and @xmath115 are the coordinates on the detection plane .
@xmath116 is the two - photon amplitude function on the double - slit plane and @xmath117 is the transmission function for a slit of width @xmath118 , so that @xmath119 if @xmath120 or @xmath121 and is equal to 1 otherwise .
the integrals of eq .
( [ sincpm ] ) were fit by the @xmath122 functions and the parameters @xmath123 and @xmath124 were extracted .
the resulting phenomenological parameters , @xmath123 and @xmath124 , were used as input for the simulated maps presented in fig . [ entangled_diffraction](c ) and fig . [ entangled_diffraction](f ) .
the simulation uses @xmath122 functions on @xmath110 and @xmath111 with effective diffraction coefficients @xmath123 and @xmath124 as an envelope for the coincidence probability maps of eq .
( [ pdbc ] ) and eq .
( [ psbc ] ) . for the dbc case , @xmath125 and @xmath126 , where @xmath127 is the expected diffraction coefficient for a model that considers the amplitude constant over the length of a slit . for same branch coincidences , @xmath128 was found by tracing over one photon of the pair and then proceeding with the numerical integration .
we found the results to be in good agreement with the experimental maps .
a similar asymmetry is reported in @xcite for a photon pairs passing through a blazed grating . there ,
also , the asymmetry is related to the degree of entanglement between the photons . ) and eq .
( [ sincpm ] ) .
the lower row , from * ( d ) * to * ( f ) * , corresponds respectively to the coincidences between the h and v branches ( d2d3 ) and to the same map corrected by subtracting the uncorrelated contribution ( d2d3-d2*d3 ) , as well as a simulated dbc map , using eq .
( [ pdbc ] ) and eq .
( [ sincpm ] ) .
the simulation parameters used are , as before , @xmath104 and @xmath105 . for a measurement time of per point
, the maximum in map * ( a ) * corresponds to 75 coincidence counts and in map * ( d ) * to 117 coincidence counts . ] in the near - separable case , @xmath116 was indeed constant over the aperture function @xmath129 . thus , we found by numerical integration of the equations represented in eq .
( [ sincpm ] ) that @xmath130 .
this was expected , since the photons have almost no entanglement and , therefore , should behave as four independent photons diffracting through the slit aperture .
it can be seen in fig .
[ diffraction_separable ] that the envelopes for both sbc and dbc are very similar , with minor differences attributed in part to a small degree of entanglement present and also to image rendering artifacts .
in this work we demonstrated experimental control of quantum correlations in a four - photon path state comprised of two pairs of photons . our scheme adopted a high - efficiency type - ii parametric down - conversion source coupled to a double slit by a variable linear optical setup , in order to obtain spatially encoded qubits . by altering experimental parameters ,
we have generated pairs with very high and very low degree of entanglement , and shown that the quantum correlations influence the two - photon interference pattern .
that makes it possible to use a map of coincidence measurements for characterizing the four - photon path state .
moreover , we have shown that , in the case of type - ii multi - pair generation , maps for coincidences between photons with the same polarization , as well as between photons with orthogonal polarization , can be used to obtain a full understanding of the phenomenon without having to resort to fourfold coincidence detection .
finally , we observed that the diffraction pattern is also affected by the degree of transverse quantum correlation between the two photons of a pair at the double - slit plane .
we were able to present a semi - quantitative approach that resulted in simulated maps in good agreement to the experimental results . a more detailed study aiming for a closed theoretical description of how entanglement plays a part in distorting the diffraction patterns observed on coincidence measurements is under way . currently , the main impediment for more efficient multi - pair experiments in quantum imaging is the fact that the area integrated by the detector is very small .
the fast improvement on the field of ccd detectors , both in readout time and in sensitivity , opens the possibility for an experiment with four or more multi - pixel detectors with largely increased genuine multi - pair detection signal @xcite .
another direction of improvement is the expansion of the dimension of the logical basis from qubits to qudits .
this can be implemented by increasing the number of slits and adjusting the quantum correlation control setup in an appropriate way . with a larger space ,
many interesting options become available , such as non - locality @xcite and contextuality @xcite tests .
the authors would like to thank cnpq , capes and fapemig brazilian funding agencies , as well as the inct - iq ( instituto nacional de cincia e tecnologia para informao quntica ) for the financial support .
l . would like to thank prof .
nivaldo speziali for providing the double slits used in this work .
we acknowledge support from italian - brazilian ( cnr - cnpq ) contract ( quantum information in a high dimension hilbert space ) . | control of spatial quantum correlations in bi - photons is one of the fundamental principles of quantum imaging . up to now
, experiments have been restricted to controlling the state of a single bi - photon , by using linear optical elements . in this work
we demonstrate experimental control of quantum correlations in a four - photon state comprised of two pairs of photons .
our scheme is based on a high - efficiency parametric down - conversion source coupled to a double slit by a variable linear optical setup , in order to obtain spatially encoded qubits . both entangled and separable pairs have been obtained , by altering experimental parameters .
we show how the correlations influence both the interference and diffraction on the double slit .
10 g. jaeger , m. horne , and a. shimony , `` complementarity of one - particle and two - particle interference , '' phys .
rev .
a * 48 * , 10231027 ( 1993 ) .
m. a. horne and a. zeilinger , `` two - particle interferometry , '' phys .
rev .
lett . *
62 * , 22092212 ( 1989 ) .
p. h. souto ribeiro , s. pdua , j. c. m. da silva , and g. a. barbosa , `` controlling the degree of visibility of young s fringes with photon coincidence measurements , '' phys
. rev .
a * 49 * , 41764179 ( 1994 ) .
t. pittman , d. strekalov , d. klyshko , m. rubin , a. sergienko , and y. shih , `` two - photon geometric optics , '' phys .
rev .
a * 53 * , 28042815 ( 1996 ) .
a. abouraddy , b. saleh , a. sergienko , and m. teich , `` role of entanglement in two - photon imaging , '' phys .
rev .
lett . * 87 * , 123602 ( 2001 ) .
b. saleh , a. abouraddy , a. sergienko , and m. teich , `` duality between partial coherence and partial entanglement , '' phys .
rev .
a * 62 * , 043816 ( 2000 ) .
i. santos , l. neves , g. lima , c. h. monken , and s. pdua , `` generation and detection of magnified images via illumination by entangled photon pairs , '' phys .
rev .
a * 72 * , 033802 ( 2005 ) .
g. brida , l. caspani , a. gatti , m. genovese , a. meda , and i. r. berchera , `` measurement of sub - shot - noise spatial correlations without background subtraction , '' phys .
rev .
lett . * 102 * , 213602 ( 2009 ) .
g. brida , m. genovese , and i. r. berchera , `` experimental realization of sub - shot - noise quantum imaging , '' nature photonics * 4 * , 227230 ( 2010 ) .
s. p. walborn , m. o. terra cunha , s. pdua , and c. h. monken , `` double - slit quantum eraser , '' phys .
rev .
a * 65 * , 033818 ( 2002 ) .
e. j. s. fonseca , c. h. monken , and s. pdua , `` measurement of the de broglie wavelength of a multiphoton wave packet , '' phys .
. lett . *
82 * , 28682871 ( 1999 ) .
e. j. s. fonseca , z. paulini , p. nussenzveig , c. h. monken , and s. pdua , `` nonlocal de broglie wavelength of a two - particle system , '' phys .
rev .
a * 63 * , 043819 ( 2001 ) .
g. lima , l. neves , i. santos , j. a. gmez , c. saavedra , and s. pdua , `` propagation of spatially entangled qudits through free space , '' phys
. rev .
a * 73 * , 032340 ( 2006 ) .
l. neves , s. pdua , and c. saavedra , `` controlled generation of maximally entangled qudits using twin photons , '' phys .
rev .
a * 69 * , 042305 ( 2004 ) .
l. neves , g. lima , j. a. gmez , c. monken , c. saavedra , and s. pdua , `` generation of entangled states of qudits using twin photons , '' phys .
. lett . *
94 * , 100501 ( 2005 ) .
l. neves , g. lima , e. fonseca , l. davidovich , and s. pdua , `` characterizing entanglement in qubits created with spatially correlated twin photons , '' phys .
rev .
a * 76 * , 032314 ( 2007 ) .
g. taguchi , t. dougakiuchi , n. yoshimoto , k. kasai , m. iinuma , h. f. hofmann , and y. kadoya , `` measurement and control of spatial qubits generated by passing photons through double slits , '' phys .
rev .
a * 78 * , 012307 ( 2008 ) .
w. peeters , j. renema , and m. v. exter , `` engineering of two - photon spatial quantum correlations behind a double slit , '' phys .
rev .
a * 79 * , 043817 ( 2009 ) .
j. armstrong , n. bloembergen , j. ducuing , and p. pershan , `` interactions between light waves in a nonlinear dielectric , '' physical review ( 1962 ) .
f. wong , m. a. albota , f. koenig , c. e. kuklewicz , e. j. mason , g. messin , and j. h. shapiro , `` high - flux entanglement sources using periodically - poled nonlinear crystals , '' nonlinear optics : materials , fundamentals and applications ( 2002 ) , paper fc8 ( 1 ) . c. kuklewicz , m. fiorentino , g. messin , f. wong , and j. shapiro , `` high - flux source of polarization - entangled photons from a periodically poled i@xmath0 parametric down - converter , '' physical review a * 69 * , 13807 ( 2004 ) .
z .- y . j. ou , _ multi - photon quantum interference _
( springer , 2007 ) , 1st ed . w. wootters , `` entanglement of formation of an arbitrary state of two qubits , '' phys . rev .
lett .
* 80 * , 22452248 ( 1998 ) . c. monken , p. ribeiro , and s. pdua , `` transfer of angular spectrum and image formation in spontaneous parametric down - conversion , '' phys .
rev .
a * 57 * , 31233126 ( 1998 ) .
o. cosme , a. delgado , g. lima , c. h. monken , and s. pdua , `` controlling the transverse correlation in qpm parametric down - conversion , '' arxiv:0906.4734v1 [ quant - ph ] .
j. w. goodman , _ introduction to fourier optics _ ( roberts & company publishers , 2004 ) , 3rd ed .
l. mandel and e. wolf , _ optical coherence and quantum optics _ ( cambridge university press , 1995 ) .
a. abouraddy , m. nasr , b. saleh , a. sergienko , and m. teich , `` demonstration of the complementarity of one- and two - photon interference , '' physical review a * 63 * , 063803 ( 2001 ) .
m. ostermeyer , d. puhlmann , and d. korn , `` quantum diffraction of biphotons at a blazed grating , '' journal of the optical society of america b * 26 * , 23472356 ( 2009 ) .
d. collins , n. gisin , n. linden , s. massar , and s. popescu , `` bell inequalities for arbitrarily high - dimensional systems , '' phys .
rev .
lett .
* 88 * , 040404 ( 2002 ) .
r. t. thew , a. acn , h. zbinden , and n. gisin , `` bell - type test of energy - time entangled qutrits , '' phys .
. lett . * 93 * , 010503 ( 2004 ) .
t. vrtesi , s. pironio , and n. brunner `` closing the detection loophole in bell experiments using qudits , '' phys .
rev .
lett . * 104 * , 060401 ( 2010 ) .
a. cabello and m. terra cunha , `` proposal of a two - qutrit contextuality test free of the finite precision and compatibility loopholes , '' arxiv:1009.2330v1 [ quant - ph ] ( 2010 ) . |
the charged particle multiplicity distribution is one of the most basic observables in high energy collisions .
although there is an abundance of experimental results , see e.g. , @xcite , on the theory side this problem is poorly understood . the multiplicity distribution measured in proton - proton ( @xmath0 ) collisions is often parameterized by the negative binomial ( nb ) distribution grosseoetringhaus:2009kz , ansorge:1988kn , breakstone:1983ns , alexopoulos:1998bi , which is characterized by two parameters : the mean number of particles @xmath2 , and @xmath3 , which measures the deviation from the poisson distribution . $ ] , which goes to poisson if @xmath4 ( at fixed @xmath5 ) . ]
nb distribution works well , with certain limitations @xcite , for a broad range of energies and in total and limited phase - space rapidity bins . for completeness
we add that @xmath3 is a decreasing function of energy .
interestingly , similar experimental observations were made in electron - positron ( @xmath1 ) collisions , see e.g. , grosseoetringhaus:2009kz , buskulic:1995xz .
nb works relatively well for total and limited phase - space bins in rapidity and @xmath3 decreases with energy .
there are many similarities between @xmath0 and @xmath1 but there are also important differences . at the same @xmath6 ,
the mean number of particles and @xmath3 are significantly larger in @xmath1 than in @xmath0 .
gev in @xmath0 collisions @xmath7 ( full phase - space ) in comparison to @xmath8 in @xmath1 at @xmath9 gev ( or @xmath10 when extrapolated to @xmath11 gev ) . ] as pointed out in refs .
@xcite some differences between @xmath0 and @xmath1 can be easily understood . in @xmath12 collisions a large fraction of available energy , given by @xmath6 ,
is carried away by two leading protons and is not available for particle production .
this explains the larger mean multiplicity in @xmath1 , where the leading proton effect is not present . as noticed in ref .
@xcite , this leads to the striking relation between total ( full phase - space ) mean number of particles in @xmath0 and @xmath1 interactions @xcite , see also batista:1998ry , sarkisyan:2010 kb : @xmath13that is , the mean number of particles in @xmath0 at a given @xmath6 is given by the mean number of particles in @xmath1 at @xmath14 , plus two leading protons .
this formula works surprisingly well from @xmath15 to @xmath16 gev ( see fig .
@xmath17 in ref .
@xcite ) and it remains to be verified at the lhc energy .
more recently , certain similarities between @xmath1 and ultra - relativistic au+au collisions at rhic were reported @xcite .
( [ mean_1 ] ) is suggestive of a universal mechanism of particle production in both systems , controlled mainly by the actual energy deposited into particle creation @xcite . in @xmath1
all available energy is consumed by produced particles , whereas in @xmath0 the _ effective _ energy coming into particle production is given by given by @xmath18 , where @xmath19 and @xmath20 is a proton mass . ]
@xmath21where @xmath22 and @xmath23 are the incoming and the leading proton momenta , respectively .
@xmath24 is a fraction of the longitudinal momentum carried by a leading proton , @xmath25 , and @xmath26 .
it this letter we show that eq .
( [ mean_1 ] ) can be extended to the whole multiplicity distribution .
we demonstrate that nb with small values of @xmath3 in @xmath0 results from nb in @xmath1 once the effective energy in @xmath0 is taken into account .
the problem of multiplicity distribution is naturally more complicated than the mean number of particles . to proceed we need to specify the leading proton @xmath27 distribution .
we choose the beta distribution @xmath28it is supported by rather limited experimental results chapman:1973kn , basile:1983 kb , however it seems a natural first choice . having ( [ f_x ] )
we obtain@xmath29which is the average momentum taken by a leading proton .
next we would like to clarify how @xmath30 is related to eq .
( [ mean_1 ] ) .
we obtain s of two leading protons are uncorrelated.]@xmath31taking@xmath32where @xmath33 @xcite we arrive at @xmath34it means that @xmath35 from eq .
( [ mean_1 ] ) is not related to @xmath36 , as naively expected , but to @xmath37 .
the latter can be calculated analytically leading to the following equation@xmath38which constrains the possible parameters of the beta distribution .
assuming @xmath39 @xcite , see eq .
( [ x_m ] ) , we obtain @xmath40 and @xmath41 , which fully determines @xmath42 .
the multiplicity distribution in @xmath0 collisions , @xmath43 , is related to the multiplicity distribution in @xmath1 interactions , @xmath44 , as @xmath45where @xmath46 .
this equation is a straightforward generalization of eq .
( [ mean_2 ] ) . instead of directly calculating the integral ( [ mean_2 ] )
we performed our calculations as follows .
first we sampled @xmath47 and @xmath48 of two leading protons from the beta distribution , @xmath30 , with @xmath49 and @xmath50 .
this allows to calculate the effective energy gev so that at least two pions can be produced .
] , @xmath51 , available for particle production in @xmath0 .
our choice of @xmath52 and @xmath53 ensures that eq .
( [ mean_1 ] ) is satisfied with the right coefficient @xmath35 .
next we sampled the number of particles from the multiplicity distribution measured in @xmath1 collisions at @xmath54 .
we take nb with the mean given by eq .
( [ n_ee ] ) , where @xmath55 @xmath56 , @xmath57 and @xmath58 where @xmath59 and @xmath60 grosseoetringhaus:2009kz .
is rounded to the nearest integer value .
nb with a negative integer @xmath3 becomes binomial distribution with the number of trials @xmath61 and the bernoulli success probability @xmath62 .
we also checked that the poisson distribution for @xmath63 leads to practically the same results . ] on top of that we add two particles corresponding to the leading protons .
we performed our calculations at @xmath64 @xmath65 @xmath66 @xmath67 @xmath68 and @xmath69 gev .
the results are shown in fig .
[ fig : nbd ] , where the calculated multiplicity distributions in @xmath0 collisions ( open symbols ) are compared with nb fits .
it is worth mentioning that eq .
( [ mean_1 ] ) is satisfied by construction so our multiplicity distributions have the correct mean values , see tab .
[ tab:2 ] .
.calculated mean number of charged particles in @xmath0 collisions compared with the experimental data . [ cols="^,^,^",options="header " , ]
several comments are in order .
\(i ) we do not offer any explanation of multiplicity distributions in @xmath70 collisions .
our goal was to demonstrate that the problem of multiplicity distributions in @xmath0 could be reduced to @xmath1 once the fluctuating energy carried away by two leading protons is taken into account .
\(ii ) in this paper we focused on the total phase - space multiplicity distributions .
it is plausible that , as assumed in this letter , the total number of particles is determined ( mostly ) by the amount of available energy .
this is not obvious for limited phase - space bins since the distribution of particles in transverse momentum or rapidity may be modified by some nontrivial dynamics .
this problem is under investigation .
\(iii ) the main uncertainty of our approach is the leading proton @xmath27 distribution given in eq .
( [ f_x ] ) .
this form is partly supported by existing data , however at rather limited energies and ranges of @xmath27 .
thus it should be considered more as an educated guess .
in addition , we assumed that @xmath30 is energy independent , which is not proven experimentally .
a mild energy dependence is indicated by theoretical studies of batista:1998ry .
\(iv ) as seen in fig .
[ fig : nbd ] the nb fits overestimate calculated multiplicity distributions for higher values of @xmath71 .
interestingly , similar trend is seen in experimental data , see , e.g. , fig .
@xmath72 in ref .
grosseoetringhaus:2009kz and figs .
@xmath73 in ref .
@xcite .
\(v ) in this paper we extrapolate @xmath74 into higher energies using 3nlo qcd result @xcite , which is almost identical to nlo qcd fit and eq . ( [ n_ee ] ) . for @xmath75
we assume it is a linear function of @xmath76 up to the lhc energies .
\(vi ) it would be interesting to extend our study on different colliding systems including proton - nucleus and nucleus - nucleus collisions akindinov:2007rr .
in conclusion , we argued that the multiplicity distribution in @xmath0 collisions is directly related to the multiplicity distribution in @xmath70 interactions , once the leading proton effect in @xmath0 is properly accounted . in @xmath0
a large fraction of initial energy , roughly @xmath77 on average , is carried away by two leading protons and is not available for particle production .
this component fluctuates from event to event , which results in a broader multiplicity distribution in @xmath0 than in @xmath1 .
we provide a new argument in favor of a common mechanism of particle production in both systems , which is mainly controlled by the amount of energy available for particle production .
i thank andrzej bialas for very useful discussions . supported by the ministry of science and higher education ( mnisw ) , by founding from the foundation for polish science , and by the national science centre ,
dec-2012/06/a / st2/00390 .
j. f. grosse - oetringhaus and k. reygers , j. phys .
g * 37 * , 083001 ( 2010 ) [ arxiv:0912.0023 [ hep - ex ] ] .
r. e. ansorge _ et al . _
[ ua5 collaboration ] , z. phys .
c * 43 * , 357 ( 1989 ) .
a. breakstone _ et al . _
[ ames - bologna - cern - dortmund - heidelberg - warsaw collaboration ] , phys .
d * 30 * , 528 ( 1984 ) .
t. alexopoulos , e. w. anderson , n. n. biswas , a. bujak , d. d. carmony , a. r. erwin , l. j. gutay and a. s. hirsch _ et al .
_ , phys . lett .
b * 435 * , 453 ( 1998 ) .
r. szwed , g. wrochna and a. k. wroblewski , acta phys .
b * 19 * , 763 ( 1988 ) .
d. buskulic _ et al . _
[ aleph collaboration ] , z. phys .
c * 69 * , 15 ( 1995 ) .
m. basile , g. cara romeo , l. cifarelli , a. contin , g. dali , p. di cesare , b. esposito and p. giusti _ et al .
_ , phys .
b * 92 * , 367 ( 1980 ) .
m. basile , g. cara romeo , l. cifarelli , a. contin , g. dali , p. di cesare , b. esposito and p. giusti _ et al .
_ , phys .
b * 95 * , 311 ( 1980 ) .
m. basile , g. bonvicini , g. cara romeo , l. cifarelli , a. contin , m. curatolo , g. dali and c. del papa _ et al .
_ , nuovo cim .
a * 73 * , 329 ( 1983 ) . m. batista and r. j. m. covolan , phys . rev .
d * 59 * , 054006 ( 1999 ) [ hep - ph/9811425 ] .
e. k. g. sarkisyan and a. s. sakharov , eur .
j. c * 70 * , 533 ( 2010 ) [ arxiv:1004.4390 [ hep - ph ] ] .
b. b. back _ et al .
_ [ phobos collaboration ] , phys .
c * 74 * , 021902 ( 2006 ) .
j. w. chapman , j. w. cooper , 1 , n. green , a. a. seidl , j. c. van der velde , c. bromberg , d. h. cohen and t. ferbel _ et al .
_ , phys .
lett . * 32 * , 257 ( 1974 ) .
m. basile , g. bonvicini , g. cara romeo , l. cifarelli , a. contin , m. curatolo , g. dali and c. del papa _ et al .
_ , lett .
nuovo cim . * 38 * , 359 ( 1983 ) .
p. ghosh , phys .
d * 85 * , 054017 ( 2012 ) [ arxiv:1202.4221 [ hep - ph ] ] .
i. m. dremin and j. w. gary , phys .
* 349 * , 301 ( 2001 ) [ hep - ph/0004215 ] .
a. akindinov , a. alici , p. antonioli , s. arcelli , m. basile , g. cara romero , m. chumakov and l. cifarelli _ et al .
_ , eur .
j. c * 50 * , 341 ( 2007 ) [ arxiv:0709.1664 [ hep - ph ] ] . | it is argued that the multiplicity distribution in proton - proton ( @xmath0 ) collisions , which is often parameterized by the negative binomial distribution , results from the multiplicity distribution measured in electron - positron ( @xmath1 ) collisions , once the fluctuating energy carried by two leading protons in @xmath0 is taken into account . |
throughout this paper , @xmath7 , @xmath8 , and @xmath9 will denote the set of positive integers , the set of nonnegative integers , and the set of prime numbers , respectively . for any function @xmath10 , we will write @xmath11 and @xmath12 for all @xmath13 .
the letter @xmath14 will always denote a prime number . for any @xmath15
, @xmath16 will denote the unique nonnegative integer @xmath17 such that @xmath18 and @xmath19 .
finally , in the canonical prime factorization @xmath20 of a positive integer , it is understood that , for all distinct @xmath21 , we have @xmath22 , @xmath23 , and @xmath24 . the well - known euler @xmath25 function is defined to be the number of positive integers less than or equal to @xmath26 that are relatively prime to @xmath26 . for each @xmath27 , the schemmel totient function @xmath28 is defined as the number of positive integers @xmath29 such that @xmath30 for all @xmath31 @xcite . in particular , @xmath32 . for no reason other than a desire to avoid cumbersome notation and the possibility of dealing with undefined objects such as @xmath33 , we will define @xmath34 to be @xmath3 for all positive integers @xmath5 . for any integer @xmath35 ,
let @xmath36 be the smallest prime number that divides @xmath26 .
schemmel @xcite showed that , for any positive integer @xmath5 , @xmath0 is multiplicative .
thus , @xmath37 .
furthermore , for @xmath35 , @xmath38 letting @xmath20 be the canonical prime factorization of @xmath26 , we may rewrite the above formula as @xmath39 for @xmath35 . in 1929 ,
s. s. pillai introduced a function that counts the number of iterations of the euler @xmath25 function needed to reach @xmath4 @xcite . in the following section ,
we generalize pillai s function via the schemmel totient functions .
then , in the third section , we generalize the concept of perfect totient numbers with the introduction , for each positive integer @xmath5 , of a function @xmath6 , which sums the first @xmath1 iterates of @xmath0 .
we record the following propositions , which follow immediately from , for later use .
[ res1.1 ] for @xmath40 , if @xmath41 , then @xmath42 .
repeatedly applying proposition [ res1.1 ] , we find [ res1.2 ] for @xmath43 , if @xmath41 , then @xmath44 .
[ res1.3 ] for @xmath45 , if @xmath5 is even and @xmath26 is odd , then either @xmath46 or @xmath28 is odd .
in addition , the following theorem is now quite easy to prove .
[ thm1.1 ] for any prime number @xmath14 and positive integer @xmath47 , @xmath48 if @xmath49 , it follows from the multiplicativity of @xmath50 that @xmath51 .
if @xmath52 , then we have @xmath53 notice that , for any positive integers @xmath5 and @xmath26 with @xmath35 , we have @xmath54 and @xmath55 . it is easy to see that , by starting with a positive integer @xmath26 and iterating the function @xmath0 a finite number of times , we must eventually reach either @xmath3 or @xmath4 .
more precisely , there exists a positive integer @xmath17 such that @xmath56 .
this leads us to the following definitions . for all @xmath45 , let @xmath57 denote the least positive integer @xmath17 such that @xmath56
furthermore , we define the function @xmath2 by @xmath58 though the functions @xmath2 only take values @xmath3 and @xmath4 , they prove to be surprisingly interesting .
for example , we can show that , for each positive integer @xmath5 , @xmath2 is a completely multiplicative function .
first , however , we will need some definitions and preliminary results . for @xmath59 , we define the following sets : @xmath60 @xmath61 @xmath62 we define @xmath63 to be the unique set of positive integers defined by the following criteria : * @xmath64 . *
if @xmath14 is prime , then @xmath65 if and only if @xmath66 .
* if @xmath47 is composite , then @xmath67
if and only if there exist @xmath68 such that @xmath69 and @xmath70 . [ lem1.1 ] let @xmath71 .
if all the prime divisors of @xmath17 are in @xmath63 , then all the positive divisors of @xmath17 ( including @xmath17 ) are in @xmath63 .
conversely , if @xmath72 , then every positive divisor of @xmath17 is an element of @xmath63 .
first , suppose that all the prime divisors of @xmath17 are in @xmath63 , and let @xmath73 be a positive divisor of @xmath17 .
then all the prime divisors of @xmath73 are in @xmath63 .
let @xmath74 be the canonical prime factorization of @xmath73 .
as @xmath75 , the third defining criterion of @xmath63 tells us that @xmath76 . then , by the same token , @xmath77 .
eventually , we find that @xmath78 . as @xmath79
, we have @xmath80 . repeatedly using the third criterion
, we can keep multiplying by primes until we find that @xmath81 .
this completes the first part of the proof .
now we will prove that if @xmath72 , then every positive divisor of @xmath17 is an element of @xmath63 .
the proof is trivial if @xmath17 is prime , so suppose @xmath17 is composite .
we will induct on @xmath82 , the number of prime divisors ( counting multiplicities ) of @xmath17 .
if @xmath83 , then , by the third defining criterion of @xmath63 , the prime divisors of @xmath17 must be elements of @xmath63 .
therefore , if @xmath83 , we are done . now , suppose the result holds whenever @xmath84 , where @xmath85 is a positive integer .
consider the case in which @xmath86 . by the third defining criterion of @xmath63 , we can write @xmath87 , where @xmath88 and @xmath89 . by the induction hypothesis , all of the positive divisors of @xmath90 and all of the positive divisors of @xmath91 are in @xmath63 . therefore , all of the prime divisors of @xmath17 are in @xmath63 . by the first part of the proof ,
we conclude that all of the positive divisors of @xmath17 are in @xmath63 .
[ thm1.2 ] if @xmath5 is a positive integer , then @xmath92 .
fix @xmath93 .
let @xmath94 be a positive integer such that , for all @xmath95 , either @xmath96 and @xmath72 or @xmath97 and @xmath98
. we will show that @xmath99 if and only if @xmath100 .
first , we must show that if @xmath95 , then @xmath96 if and only if @xmath101 .
suppose , for the sake of finding a contradiction , that @xmath101 and @xmath97 .
as @xmath97 , we have that @xmath102 and @xmath98 .
lemma [ lem1.1 ] then guarantees that there exists a prime @xmath103 such that @xmath104 and @xmath105 . as @xmath105 ,
the second defining criterion of @xmath63 implies that @xmath106 .
we know that @xmath107 because , otherwise , @xmath108 , implying that @xmath109
. therefore , @xmath110 and @xmath106 . by the induction hypothesis , @xmath111 .
therefore , there exists some @xmath112 such that @xmath113 . because @xmath104 , proposition [ res1.1 ] implies that @xmath114 .
thus , @xmath115 , which implies that @xmath116 .
this is a contradiction .
now suppose , so that we may again search for a contradiction , that @xmath116 and @xmath96 .
@xmath116 implies that @xmath102 , and @xmath96 implies ( by the induction hypothesis ) that @xmath72 . by lemma [ lem1.1 ] ,
all positive divisors of @xmath17 are elements of @xmath63 .
let @xmath117 be the canonical prime factorization of @xmath17 .
then , by , @xmath118 if @xmath119 , then @xmath120 , which mean that @xmath121 .
as @xmath122 , we have contradicted @xmath96 .
therefore , @xmath123 , so @xmath124 . for each @xmath125
, @xmath126 is a positive divisor of @xmath17 , so @xmath127 .
the second criterion defining @xmath63 then implies that @xmath128 , so all positive divisors ( and , specifically , all prime divisors ) of @xmath129 are in @xmath63 .
this implies that all prime divisors of @xmath130 are elements of @xmath63 , so lemma [ lem1.1 ] guarantees that @xmath131 .
however , we have shown that @xmath132 , so @xmath133 . by the induction hypothesis
, we have @xmath101 , a contradiction .
thus , we have established that if @xmath95 , then @xmath96 if and only if @xmath101 . we are now ready to establish that @xmath99 if and only if @xmath100 .
suppose that @xmath99 and @xmath134 .
we know that @xmath135 because , otherwise , @xmath136 , implying that @xmath137 and contradicting @xmath99 .
if @xmath94 is prime , then @xmath99 implies that @xmath138 . then @xmath139 . as @xmath140 ( we assume here and in the rest of the proof that @xmath141 is large enough so that the notation @xmath142 makes sense as we have defined it , but the argument is valid in any case ) , it follows from the preceding argument that @xmath143 . continuing this pattern , we eventually find that @xmath144 , so @xmath145 . by the induction hypothesis , @xmath146 .
however , by the second criterion defining @xmath63 , the primality of @xmath94 then implies that @xmath100 , a contradiction .
thus , @xmath94 must be composite .
we assumed that @xmath134 , so lemma [ lem1.1 ] guarantees the existence of a prime @xmath105 such that @xmath147 .
as @xmath94 is composite , @xmath148 .
the induction hypothesis then implies that @xmath149 , so @xmath150 .
however , this contradicts @xmath99 , so we have shows that if @xmath99 , then @xmath100 .
suppose , on the other hand , that @xmath151 and @xmath100 . again , we begin by assuming @xmath94 is prime .
then , because @xmath100 , we must have @xmath146 . therefore , by the induction hypothesis and the fact that @xmath152 , it follows that @xmath145 .
now , @xmath151 , so we must have @xmath153 .
therefore , @xmath154 .
however , as @xmath155 , it follows that @xmath156 .
again , we continue this pattern until we eventually find that @xmath157 , which means that @xmath158 .
this is a contradiction , and we conclude that @xmath94 must be composite . from @xmath100 and lemma [ lem1.1 ] ,
we conclude that all of the prime divisors of @xmath94 are elements of @xmath63 .
furthermore , as @xmath94 is composite , all of the prime divisors of @xmath94 are elements of @xmath159
. then , by the induction hypothesis , all of the prime divisors of @xmath94 are in the set @xmath160 .
this implies that none of the prime divisors of @xmath94 are in @xmath161 , so @xmath99 .
this is a contradiction , and the induction step of the proof is finally complete .
all that is left to check is the base case .
however , the base case is trivial because @xmath162 and @xmath64 .
we may now use the sets @xmath160 and @xmath63 interchangeably .
in addition , part of the above proof gives rise to the following corollary .
[ corol ] let @xmath163 .
then @xmath164 if and only if @xmath165 .
the proof follows from the argument in the above proof that @xmath166 if and only if @xmath165 whenever @xmath167 . as we now know
that we can make @xmath94 as large as we need , it follows that @xmath166 if and only if @xmath165 .
then @xmath168 if and only if @xmath166 , @xmath169 if and only if @xmath168 , and , in general , @xmath170 if and only if @xmath171 @xmath172 .
the desired result follows immediately .
[ cor1.1 ] let @xmath173 .
then @xmath174 if and only if @xmath165 .
it is clear that @xmath174 if and only if @xmath175 .
therefore , the proof follows immediately from setting @xmath176 in corollary [ corol ] .
notice that , for a given positive integer @xmath5 , corollary [ cor1.1 ] , along with theorem [ thm1.2 ] and the defining criteria of @xmath63 , provides a simple way to construct the set of all positive integers @xmath47 that satisfy @xmath177 .
corollary [ cor1.1 ] also expedites the proof of the following theorem .
[ thm1.3 ] the function @xmath178 is completely multiplicative for all @xmath27 .
choose some @xmath179 .
first , suppose @xmath180 . by corollary [ cor1.1 ] ,
therefore , there exists @xmath150 such that @xmath182 .
this implies that @xmath183 , so @xmath184 .
thus , @xmath185 .
a similar argument shows that @xmath185 if @xmath186 . now , suppose that @xmath187 . then corollary [ cor1.1 ] ensures that @xmath188 .
therefore , @xmath189 , so @xmath190 .
as the function @xmath2 can only take values @xmath3 and @xmath4 , the proof is complete . in concluding this section
, we note that if @xmath191 is composite , then it is impossible for any integer greater than @xmath4 to be in @xmath160
. therefore , whenever @xmath191 is composite , we have @xmath192 and @xmath193 for all integers @xmath35 .
a perfect totient number is defined @xcite to be a positive integer @xmath35 that satisfies ( using our previous notation ) @xmath194 in the following definitions , we generalize the concept of perfect totient numbers .
we also borrow some other traditional terminology related to perfect numbers .
let @xmath5 be a positive integer .
we define the arithmetic function @xmath6 by @xmath195 and @xmath196 for all integers @xmath35 . if @xmath197 , we say that @xmath26 is @xmath6__-deficient__. if @xmath198 , we say that @xmath26 is @xmath6__-perfect__. if @xmath199 , we say that @xmath26 is @xmath6__-abundant__. finally , in the case when @xmath200 , we say that @xmath26 is @xmath6__-stagnant__. we now present a series of theorems related to these definitions . [ thm2.1 ]
if @xmath201 is odd , then all positive integers are @xmath6-deficient .
let @xmath201 be an odd integer , and let @xmath26 be any positive integer . if @xmath202 or @xmath203 , then @xmath26 is @xmath6-stagnant . _ a fortiori _
, @xmath26 is @xmath6-deficient . if @xmath204 , then @xmath205 is even and @xmath206 ( by proposition [ res1.1 ] ) .
thus , @xmath207 , which implies that @xmath208 . hence , @xmath209 .
[ thm2.2 ] all positive even integers are @xmath6-deficient for all positive integers @xmath5 .
the proof is trivial for @xmath201 because , in that case , any positive even integer is clearly @xmath6-stagnant . for @xmath210 ,
we use the fact that all totient numbers greater than @xmath4 are even .
therefore , @xmath211 theorem [ thm2.2 ] is nothing revolutionary , but we include it because it fits nicely with the next theorems . for the next two theorems , which are not quite as trivial as the previous two , we require the following lemma . [ lem2.1 ] if @xmath102 is an odd integer , then at least one element of the set @xmath212 is divisible by @xmath213 .
let @xmath102 be an odd integer with prime factor @xmath14 , and suppose @xmath214 and @xmath215 .
we know that @xmath216 ( mod @xmath213 ) because @xmath217 , so @xmath218 ( mod @xmath213 ) . as @xmath219 ( mod @xmath213 )
, @xmath220 must have some prime factor @xmath221 such that @xmath222 ( mod @xmath213 ) .
but then , using proposition [ res1.2 ] , @xmath223 . [ thm2.3 ] for any integer @xmath201 , all positive multiples of @xmath213 are @xmath6-deficient . if @xmath224 , then any positive multiple of @xmath213 is clearly @xmath6-stagnant .
therefore , we only need to check the case @xmath225 .
write @xmath226 .
suppose @xmath227 and let @xmath228 be the smallest element of @xmath229 . if @xmath230 for some @xmath231 , then @xmath232 .
therefore , @xmath228 must have some prime divisor @xmath233
. from theorem [ thm2.1 ] , @xmath234 .
also , by proposition [ res1.2 ] , @xmath235 and @xmath236 . by lemma [ lem2.1 ] , at least one of @xmath237 and @xmath238 must be divisible by @xmath213 .
suppose @xmath239 so that @xmath240 . by the choice of @xmath228 as the smallest element of @xmath229
, we have @xmath241 .
this implies that @xmath242 because @xmath243 . from this contradiction
, we conclude that @xmath244 , so @xmath245 . again , by the choice of @xmath228 , we have @xmath246 .
however , this implies that @xmath247 , which is a contradiction .
it follows that @xmath229 is empty .
[ thm2.4 ] if @xmath201 is a positive integer and @xmath248 , then all positive multiples of @xmath249 are @xmath6-deficient .
let @xmath5 be a positive integer other than @xmath4 or @xmath250 , and let @xmath26 be a multiple of @xmath249 . if @xmath251 , then @xmath26 is @xmath6-stagnant . if @xmath252 , then @xmath26 is @xmath6-deficient by theorem [ thm2.1 ] . we therefore only need to check the case @xmath225 .
write @xmath253 , where @xmath254 .
we may assume that @xmath255 and @xmath214 because , otherwise , the desired result follows immediately from either theorem [ thm2.2 ] or theorem [ thm2.3 ] .
we now consider two cases .
case 1 : @xmath256 .
write @xmath257 , where @xmath258 is a positive integer not divisible by @xmath259 , @xmath213 , or @xmath249 and @xmath260 ( we use proposition [ res1.3 ] to conclude that @xmath258 is odd )
. then @xmath261 and @xmath262 . as @xmath263
, we can use theorem [ thm2.3 ] to write @xmath264 .
this completes the proof of the case when @xmath256 .
case 2 : @xmath265 . in this case , @xmath266 , so @xmath267 .
we may assume that @xmath102 because the case @xmath268 is trivial .
first , suppose that @xmath269 . in this case , @xmath270 , and , by theorem [ thm1.1 ] , @xmath271 .
then , using theorem [ thm2.3 ] , we have @xmath272 .
now suppose that @xmath215 .
by lemma [ lem2.1 ] and our assumption that @xmath214 , we have @xmath273 . using theorem [ thm1.1 ] and theorem [ thm2.3 ] again
, we have @xmath274 .
this completes the proof of all cases .
the last few theorems have dealt with @xmath6-deficient numbers , so it is natural to ask questions about @xmath6-abundant numbers .
we might wish to know the positive integers @xmath5 for which @xmath6-abundant numbers even exist .
how many @xmath6-abundant numbers exist for a given @xmath5 ?
how large can we make @xmath275 ?
theorem [ thm2.1 ] deals with these questions for the cases when @xmath5 is odd and greater than @xmath4
. also , a great deal of literature @xcite already exists concerning the case @xmath210 . in the following theorem
, we answer all of the preceding questions for the cases when @xmath5 is a positive even integer .
[ thm2.5 ] let @xmath5 be a positive even integer .
for any positive @xmath276 and @xmath277 , there exist infinitely many primes @xmath278 such that @xmath279 for all positive integers @xmath280 .
fix @xmath5 , @xmath276 , and @xmath277 to be positive real numbers , where @xmath5 is an even integer .
let @xmath281 be all the primes that divide @xmath5 , and let @xmath282 be all the primes that are less than @xmath5 and do not divide @xmath5 .
for each @xmath283 , define @xmath284 by @xmath285 write @xmath286 . by the chinese remainder theorem
, there exists a unique solution modulo @xmath287 to the system of congruences defined by @xmath288 it is easy to see that if @xmath289 is a solution to , then @xmath289 and @xmath290 are each relatively prime to every prime less than or equal to @xmath5 . by dirichlet
s theorem concerning the infinitude of primes in arithmetic progressions , there must be infinitely many primes that satisfy the system .
let @xmath278 be one such prime , and write @xmath291 as @xmath278 is relatively prime to all primes less than or equal to @xmath5 , we have @xmath292 it is well - known @xcite , that , as @xmath293 , @xmath294 for some constant @xmath295 that depends only on @xmath5 .
we find that @xmath296 therefore , we may choose @xmath278 to be large enough so that @xmath297 . with this choice of @xmath278 , we may write @xmath298 . but @xmath299 because @xmath300 .
thus , @xmath301 let @xmath280 be an integer , and , for now , assume @xmath256 . rearranging and multiplying the inequality by @xmath302 , we have @xmath303 . after further algebraic manipulation ,
we find @xmath304 .
noticing that the left - hand side of the preceding inequality is simply @xmath305 , we have @xmath279 .
this is the desired result for @xmath256 . to show that the result holds when
@xmath265 , it suffices to show that @xmath306 .
this reduces to @xmath307 , which is obviously true .
[ cor2.1 ] for any positive even integer @xmath5 , there exist infinitely many @xmath6-abundant numbers .
we conclude this section with a remark about @xmath6-perfect numbers . using _
mathematica _ , one may check that for @xmath308 , the only @xmath6-perfect number less than @xmath309 is @xmath310 , which is @xmath311-perfect .
unfortunately , this data is too scarce to make any reasonable conjecture about the nature or distribution of @xmath6-perfect numbers for positive even integers @xmath5 .
in 1943 , h. shapiro investigated a function @xmath312 , which counts the number of iterations of the @xmath25 function needed to reach @xmath259 @xcite .
shapiro showed that the function @xmath312 is additive , and he established bounds for its values . in this paper , we have not gone into much detail exploring the functions @xmath1 because they prove , in general , to be either completely uninteresting or very difficult to handle .
for example , for any integer @xmath35 , @xmath313 on the other hand , the function @xmath314 does not seem to obey any nice pattern or exhibit any sort of nice additive behavior .
there seems to be some hope in analyzing the function @xmath315 , so we make the following conjecture .
if @xmath316 is an odd integer , then @xmath317 we note that it is not difficult to prove , using lemma [ lem2.1 ] and a bit of case work , that @xmath318 for all integers @xmath319 ( with equality only at @xmath320 ) . however , as figure 1 shows , this is a very weak upper bound ( at least for relatively small @xmath47 ) .
it is tempting to think , based on the figure , that @xmath321 for all positive integers @xmath47 .
however , setting @xmath322 yields a counterexample because @xmath323 . the author has found that investigating bounds of the function @xmath315 naturally leads to a question about the infinitude of twin primes , which hints at the potential difficulty of the problem .
indeed , harrington and jones @xcite have arrived at the same conclusion while studying the function @xmath324 , and they conjecture that the values of @xmath325 can be arbitrarily large . to avoid the unpredictability of the values of the function @xmath326 , harrington and jones have restricted the domain of @xmath326 to the set @xmath327 of positive integers @xmath17 with the property that none of the numbers in the set @xmath328 has a prime factor that is congruent to @xmath4 modulo @xmath213 . with this restriction of the domain of @xmath326 ,
these two authors have established results analogous to those that shapiro gave for the function @xmath312 mentioned earlier .
in fact , we speculate that methods analogous to those that harrington and jones have used could easily generalize to allow for analogous results concerning functions @xmath329 if one is willing to use a sufficiently restricted domain of @xmath330 .
we next remark that , in theorem [ thm2.4 ] , the requirement that @xmath248 is essential . for example , write @xmath331 , @xmath332 , @xmath333 , @xmath334 , and @xmath335
. then the number @xmath336 is a @xmath337-abundant multiple of 5 .
lastly , we have not spent much effort analyzing the sizes " of the functions @xmath6 or searching for @xmath6-perfect numbers .
we might inquire about the average order or possible upper and lower bounds for @xmath6 for a general positive even integer @xmath5 .
in addition , it is natural to ask if there even are any @xmath6-perfect numbers other than @xmath310 for even positive integers @xmath5 .
dedicated to my parents marc and susan , my brother jack , and my sister juliette . also dedicated to mr .
jacob ford , who wrote a program to find values of the functions @xmath0 , @xmath1 , and @xmath2 .
ford and my father also sparked my interest in computer programming , which i used to analyze the functions @xmath6 . | we begin by introducing an interesting class of functions , known as the schemmel totient functions , that generalizes the euler totient function . for each schemmel totient function @xmath0
, we define two new functions , denoted @xmath1 and @xmath2 , that arise from iterating @xmath0 . roughly speaking ,
@xmath1 counts the number of iterations of @xmath0 needed to reach either @xmath3 or @xmath4 , and @xmath2 takes the value ( either @xmath3 or @xmath4 ) that the iteration trajectory eventually reaches .
our first major result is a proof that , for any positive integer @xmath5 , the function @xmath2 is completely multiplicative .
we then introduce an iterate summatory function , denoted @xmath6 , and define the terms @xmath6-deficient , @xmath6-perfect , and @xmath6-abundant .
we proceed to prove several results related to these definitions , culminating in a proof that , for all positive even integers @xmath5 , there are infinitely many @xmath6-abundant numbers .
many open problems arise from the introduction of these functions and terms , and we mention a few of them , as well as some numerical results .
.2 in |
heavy hadron decays can be computed using a systematic expansion in powers of @xmath6 and @xmath7 , where @xmath8 is the mass of the heavy quark and @xmath9 is the nonperturbative scale parameter of the strong interactions . in the @xmath10 limit ,
inclusive decay rates are given by free quark decay and the order @xmath7 corrections vanish @xcite .
the leading nonperturbative corrections of order @xmath11 are parameterized by two hadronic matrix elements @xcite @xmath12 and @xmath13 .
these results are now used to determine the ckm matrix elements @xmath14 and @xmath15 , using experimental data on inclusive semileptonic @xmath0 meson decays . for exclusive decays at lowest order in @xmath16 , one can compute form factors in terms of a perturbative series in @xmath17 times the isgur - wise function @xcite .
nonperturbative corrections are parameterized by hadronic matrix elements , and are of order @xmath18 . in some cases at zero recoil ,
the order @xmath16 correction vanishes @xcite . at present ,
the largest theoretical uncertainties in the computation of inclusive @xmath19 and @xmath20 decay rates arise from poor knowledge of quark masses .
the pole mass of a heavy quark is an infrared ( ir ) sensitive quantity , and is not well defined beyond perturbation theory @xcite .
this is related to the bad behavior of perturbative corrections to the inclusive decay rate when it is written in terms of the pole mass @xcite .
a simple method of avoiding problems with the quark mass is to use instead the @xmath0 mesons mass .
unfortunately , the @xmath0 meson and @xmath2 quark masses differ by order @xmath9 , and so this reintroduces a @xmath21 correction to the inclusive decay rate .
the decay rate has also been rewritten in terms of the infrared safe @xmath22 mass of the @xmath2 quark , but the uncertainties still remain sizable because the value of the @xmath22 mass is not known very accurately , and because the first few terms in the perturbation series still remain large . in this paper ,
the theoretical predictions for semileptonic @xmath0 decays are rewritten in terms of the @xmath1 meson mass rather than the @xmath2 quark mass . at the same time a modified perturbation expansion ( referred to as the upsilon expansion ) must be used , which is discussed in detail in sec .
this procedure eliminates the uncertainty due to the @xmath23 factor in the decay rates , and at the same time improves the behavior of the perturbation series .
our formulae relate measurable quantities to one another and the resulting perturbation series is free of renormalon ambiguities .
we will also see numerically that the perturbative corrections are small when the @xmath0 decay rate is written in terms of the @xmath24 mass .
we have discussed the procedure in an earlier publication @xcite , where it was applied to inclusive semileptonic @xmath0 decays . in this article , we describe the method in more detail and also explain why it leads to an improvement in the perturbation series expansion .
the method is applied to inclusive nonleptonic decays as well as to exclusive decays .
the perturbation series is better behaved for all the processes we have examined .
the semileptonic and nonleptonic decay computations can be combined to obtain new values for the semileptonic branching ratio and the average charm multiplicity in @xmath0 meson decay .
the outline of this paper is as follows . in section
ii we define the upsilon expansion . in sec .
iii we show that it is the theoretically consistent way of eliminating the @xmath2 quark mass from the combination of the perturbation theory results for @xmath0 decay rates and the @xmath24 mass
. we also comment on the role of nonperturbative effects in the theoretical expression for the @xmath24 mass .
exclusive decays are discussed in iv . in sec .
v we give the results for inclusive semileptonic and nonleptonic decay rates in the upsilon expansion , the ratio of decay rates @xmath25 , the semileptonic @xmath0 branching ratio , and charm counting in @xmath0 decay . sec .
vi contains our conclusions and comments on further applications of our results .
the results for the ope calculation of inclusive @xmath0 decay rates can be schematically written as @xmath26 , \end{aligned}\ ] ] where the precise coefficients are not shown .
@xmath27 denotes an arbitrary final state created by the @xmath28 transition at short distance ( e.g. , @xmath29 etc . )
, @xmath30 contains ckm matrix elements , @xmath31 is the @xmath2 quark decay phase space at tree level including color factors ( e.g. , @xmath32 is a factor 3 enhanced compared to @xmath33 ) , @xmath34 is the @xmath2 quark pole mass , @xmath35 is the first coefficient of the qcd @xmath36-function , and @xmath17 is the running coupling constant in the @xmath22 scheme at the scale @xmath37 .
it is convenient to expand the coefficient of a given @xmath17 order in powers of @xmath38 which is related to the brodsky lepage mackenzie ( blm ) prescription @xcite . in the perturbation expansion . in some cases ,
the entire @xmath39 contribution has not been computed , and only the blm piece is known . in cases where the entire @xmath39 term is known , the blm contribution is about ten times larger than the non - blm part . ] for nonleptonic and rare @xmath0 decays there is another large parameter in eq .
( [ schemrate ] ) , @xmath40 , which will be discussed in sec .
these large logarithms can be summed using the renormalization group equations .
the reason for introducing the variable @xmath41 is explained in the next paragraph .
schematically , the perturbative expansion of the @xmath24 mass in terms of @xmath34 is @xmath42 \epsilon^2 \nonumber\\ * & + & { \alpha_s^2\over\pi^2}\ , \big [ \big ( \ell^2 + \ell + 1 \big ) \beta_0 ^ 2 + \ldots \big ] \epsilon^3 + \ldots \bigg\ } \,,\end{aligned}\ ] ] where @xmath43 $ ] , @xmath44 , and the precise coefficients are again not shown .
note that this series is of the form @xmath45 , whereas the corrections in eq .
( [ schemrate ] ) are of order @xmath46 .
we will show in sec .
iii that to combine these two equations in a theoretically consistent manner , terms of order @xmath47 in eq .
( [ schemups ] ) should be viewed as if they were only of order @xmath48 .
for this reason , the power of @xmath49 in eq .
( [ schemups ] ) is one less than the power of @xmath17 .
it is also convenient to choose the same renormalization scale , @xmath37 , when eqs .
( [ schemrate ] ) and ( [ schemups ] ) are combined .
the prescription of counting @xmath50^n$ ] in the @xmath0 decay rate as order @xmath51 , and @xmath50^n$ ] in the @xmath24 mass as order @xmath52 will be called the upsilon expansion .
it is expected that the infrared sensitivity present separately in eqs .
( [ schemrate ] ) and ( [ schemups ] ) will cancel to all orders in perturbation theory in @xmath49 .
some arguments that support this will be given in the next section .
in the upsilon expansion eq .
( [ schemrate ] ) takes the form @xmath53 , \ ] ] where @xmath54 means that the @xmath2 decay phase space is also expanded in @xmath49 .
we emphasize that the coefficients in the series in @xmath49 contain different orders in @xmath17 .
we show in sec .
iii that this is theoretically consistent . our results in sec .
iv and v demonstrate that this expansion simultaneously eliminates the uncertainty in the theoretical predictions related to the dependence on the poorly known quark masses and also improves the behavior of the perturbation series
. the precise form of eq .
( [ schemrate ] ) depends on the decay channel under consideration .
the expression for the @xmath24 mass in terms of @xmath34 is given by @xcite , @xmath55 \epsilon^2 \\ & + & \bigg({\alpha_s\beta_0\over2\pi}\bigg)^2 \bigg ( 3\ell^2 + 9\ell + 2\zeta(3)+\frac{\pi^2}6+\frac{77}{12}\bigg ) \epsilon^3 + \ldots \bigg\ } .
\nonumber\end{aligned}\ ] ] the ellipsis denote terms of order @xmath56 , and non - blm terms of order @xmath57 . for @xmath37 of order @xmath34 , eq .
( [ upsmass ] ) shows no sign of convergence ; for @xmath58 it yields @xmath59 , using @xmath60 .
( this value of @xmath61 will be used whenever we present numerical results . )
the bad behavior of this series is unimportant ( it would be somewhat better in terms of the @xmath22 mass at a low scale ) .
the only physical question is what happens when we use eq .
( [ upsmass ] ) to predict @xmath0 decay rates in terms of @xmath62 .
finally , decays with a charm quark in the final state depend both on @xmath34 and @xmath63 .
it is convenient to express these decay rates in terms of @xmath62 and @xmath12 instead of @xmath34 and @xmath63 , using eq .
( [ upsmass ] ) and @xmath64 where @xmath65gev and @xmath66gev .
one could imagine replacing @xmath63 by the @xmath67 mass using a procedure similar to that for the @xmath2 quark . however , the nonperturbative errors in this case are not under control .
in ref . @xcite we showed that our method of combining the expansion of inclusive @xmath0 decay rates with the expansion of the @xmath24 mass to eliminate the dependence on the @xmath2 quark mass is consistent at large orders in perturbation theory in the blm approximation . in this section
we briefly repeat this argument ( sec .
a ) and show that the upsilon expansion is also consistent at first and second order in @xmath49 , including non - abelian contributions ( sec .
our arguments prove that , as far as the cancellation of infrared sensitive contributions in the perturbative expansion of the @xmath0 decay is concerned , the upsilon expansion is equivalent to the use of a short - distance mass like the @xmath22 mass .
they also explain why the upsilon expansion requires powers of @xmath17 to be counted differently in @xmath62 and @xmath0 decay rates
. we also comment on the relation between perturbative and nonperturbative effects in the @xmath24 mass ( sec.c ) .
the pole mass @xmath34 is an infrared sensitive quantity , which is not well defined beyond perturbation theory .
it can be related to an infrared safe mass , such as the @xmath22 mass @xmath68 via ( for @xmath69 ) @xcite @xmath70 this relation has terms of the form @xmath71 at high orders , leading to a renormalon ambiguity .
the borel transform of the perturbation series relating the pole mass to the @xmath22 mass is @xcite @xmath72 \bigg\ } , \ ] ] where @xmath73 is a scheme dependent constant ( @xmath74 in @xmath22 ) .
the leading renormalon ambiguity is given by the @xmath75 pole , yielding @xmath76 the @xmath17 perturbation series for the inclusive decay rate written in terms of the pole mass also has a renormalon singularity . as a result
, the inclusive decay perturbation series is badly behaved at high orders .
renormalons cancel when the inclusive decay rate is written in terms of the @xmath22 mass @xcite , rather than the pole mass . while this cancellation is present at high orders , the perturbation series written in terms of the @xmath22 mass at a renormalization scale @xmath37 of order @xmath34 still contains large corrections at low orders ( see , ref .
@xcite and some examples in sec .
v ) . on the other hand ,
the perturbation series written in terms of the @xmath22 mass at a low scale has a better behavior .
but such a choice still does not remove the quark mass uncertainty in the decay rates .
the expansion of the @xmath24 mass also becomes better behaved when written in terms of the @xmath22 mass at a low scale ( of order the bohr radius , @xmath77 ) , rather than in terms of the pole mass .
this fact is important for extracting the short distance masses of the bottom and top quarks from the @xmath24 system and from the location of the peak of the @xmath78 @xmath79 resonance in the total cross section for @xmath79 production in @xmath80 annihilation close to threshold .
one might think that the numerical results obtained in the upsilon expansion are equivalent to the result obtained by extracting a short - distance @xmath2 quark mass from the @xmath24 mass and then using it to predict @xmath0 decay rates . due to the fact that only the first few terms in the perturbative expansions are known , this is not the case because in this procedure the information on the correlations between the short - distance mass , the strong coupling constant and the renormalization scale is lost .
the upsilon expansion , which completely eliminates the intermediate step of using any quark mass takes full advantage of these correlations and leads to smaller uncertainties .
the potential between two static quarks in a color singlet state , @xmath81 , also has renormalon ambiguities .
the borel transform is given by @xcite @xmath82 the leading renormalon ambiguity of @xmath83 also comes from the @xmath75 pole , @xmath84 it follows from eqs .
( [ borelmass ] ) and ( [ borelpot ] ) that the order @xmath9 ambiguity cancels in @xmath85 @xcite .
equation ( [ borelpot ] ) also explains the mismatch between the power of @xmath17 and @xmath49 in eq .
( [ schemups ] ) , which may seem somewhat unnatural at first sight . the coefficient of @xmath86 in eq . ( [ schemups ] )
is given by the @xmath24 matrix element of the coefficient of @xmath87 in the expansion of @xmath88 in eq .
( [ borelpot ] ) about @xmath89 , multiplied by @xmath90 . for large enough @xmath91 , up to corrections suppressed by at least one power of @xmath91 , the terms in this expansion of @xmath92 are proportional to @xmath93^p \over p ! } \simeq \mu\,(2u)^n.\ ] ] this explains why in the blm approximation and in higher orders in @xmath49 , the terms in eq .
( [ schemups ] ) are generically of the form ) differ from those displayed in eq .
( [ genericexpansion ] ) because of multiple insertions of the corrections to the pure coulomb potential , @xmath94 .
the contributions arising from multiple insertions , however , only constitute corrections suppressed by at least one power of @xmath91 . ]
@xmath95 since the @xmath24 matrix element of @xmath96 .
for large @xmath91 , @xmath97 reducing the difference between the powers of @xmath17 and @xmath38 to one , thus eliminating the mismatch in the large order behaviors of eqs .
( [ schemrate ] ) and ( [ schemups ] ) .
this has to happen since @xmath62 is a physical quantity , so the renormalon ambiguities must cancel in eq .
( [ schemups ] ) between @xmath98 and the potential plus kinetic energies @xcite .
if nrqcd exists as a well defined low energy effective field theory , then the @xmath99 renormalon in the potential corresponding to the @xmath100 pole in eq .
( [ borelpot ] ) must cancel against the leading renormalon in the kinetic energy contribution to the @xmath24 mass .
while this has not been shown explicitly , it is quite conceivable that this happens .
studying higher order renormalons is less interesting as there are certainly nonperturbative contributions of order @xmath101 to @xmath62 ( e.g. , from the vacuum expectation value of @xmath102 @xcite ) , and so higher order renormalon effects are of the same order as other nonperturbative effects which are known to exist .
one can study the infrared sensitivity of feynman diagrams by introducing a fictitious infrared cutoff @xmath103 of order @xmath9 .
the infrared sensitive terms are nonanalytic in @xmath104 , such as @xmath105 or @xmath106 .
these terms arise from the low - momentum part of feynman diagrams .
this is similar to what occurs in chiral perturbation theory , where the nonanalytic terms in the quark mass ( or equivalently , the pion mass - squared ) come from the low - momentum behavior of feynman diagrams .
diagrams which are more infrared sensitive , i.e. , have contributions @xmath105 or @xmath106 for small values of @xmath91 , should have larger nonperturbative contributions .
in particular , linear infrared sensitivity , i.e. , terms of order @xmath103 , are a signal of @xmath9 effects , quadratic sensitivity , i.e. , terms of order @xmath107 are a signal of @xmath108 effects , etc . in the following
we show that the linear infrared sensitivity present separately in the expansion of the @xmath0 decay rate and in the expansion of the @xmath24 mass in terms of the @xmath2 quark mass cancel at least to order @xmath39 when they are combined using the upsilon expansion .
( at order @xmath49 this is equivalent to the cancellation of the leading renormalon ambiguity in the blm approximation , discussed previously . )
the argument is essentially a combination of those in recent papers by beneke @xcite , and by sinkovics , akhoury , and zakharov @xcite . as far as the cancellation of infrared sensitive contributions in the perturbative expansion of @xmath0 decay
is concerned , the upsilon expansion is therefore equivalent to a scheme where a short - distance mass like the @xmath22 mass is used .
the proof proceeds in two steps .
first one notes that the linear ir sensitivity cancels in inclusive decay widths of a heavy quark when it is expressed in terms of an infrared safe short distance mass ( such as the @xmath22 mass ) . to leading order in @xmath17 this
was shown in refs .
@xcite . to order @xmath109 , where non - abelian contributions first occur
, the cancellation of the linear ir sensitivity was shown only recently in ref .
this cancellation of the linear ir sensitivity is expected to hold at higher orders in @xmath17 as well , but the demonstration of this appears highly non - trivial .
the second step is to show the cancellation of the linear ir sensitivity when the @xmath24 mass is expressed in terms of an infrared safe short distance mass .
this was done by beneke @xcite , and we will repeat some of his arguments below . the reason is to explain why it is imperative to assign one less power of @xmath49 to each term in eq .
( [ upsmass ] ) than the power of @xmath17 even at low orders in @xmath49 .
let us first consider the leading contribution to the heavy quark self energy , shown in fig .
[ fig : hqselfenergy ] .
since we are only interested in the infrared behavior , we may calculate the self energy diagram in terms of the bare mass , @xmath110 , which differs from the @xmath22 mass only by ultraviolet subtractions .
the self energy is given by @xmath111 ( q^2+i\varepsilon ) } \,.\ ] ] to obtain the leading infrared contribution to the mass shift coming from the region of small loop momentum , we can set @xmath112 , where @xmath113 is the four - velocity of the heavy quark ( @xmath114 ) .
this gives @xmath115 the integration over @xmath116 is simplest by closing the contour in the upper half plane , yielding @xmath117 of course , this formula is only valid in the infrared . restricting the range of integration to @xmath118 , where @xmath103 is of order @xmath9
, it gives @xmath119 .
let us next consider the heavy quark potential .
it was shown in ref .
@xcite that the potential in momentum space is free of linear ir sensitivity .
the linear ir sensitivity of the contribution of the potential to @xmath62 arises in a momentum space calculation from taking its matrix element .
this occurs because the momentum space wave function @xmath120 does not suppress the low momentum contribution of @xmath121 . in coordinate space ,
the wave function @xmath122 is exponentially small at long distances , and the ir sensitivity of the contribution of the potential to @xmath62 arises from fourier transforming @xmath121 into coordinate space to obtain @xmath123 .
therefore , the linear infrared sensitivity of the potential energy can be isolated by restricting the range of integration in the fourier transform @xcite @xmath124 choosing @xmath103 of order @xmath9 , much smaller than @xmath125 , the @xmath126 factor can be replaced by 1 , and the ir sensitive term is a constant times @xmath127 . is essential for the argument to hold . for the @xmath1 ground state this assumption is valid . ] since this contribution to the potential is independent of @xmath128 , its @xmath24 matrix element is simply itself , @xmath129 from eq .
( [ alphaslambda ] ) it is evident that the infrared sensitive contribution in the @xmath24 matrix element of the heavy quark potential is of one lower order in @xmath17 than the @xmath24 matrix element of the heavy quark potential itself in the limit @xmath130 , @xmath131 this mismatch arises because the @xmath24 matrix element of @xmath132 is of order the inverse bohr radius @xmath133 .
consequently , it is the infrared sensitive contribution in the order @xmath109 term in the perturbative series for @xmath62 in eq .
( [ upsmass ] ) which cancels the infrared sensitive contribution of the order @xmath17 term in the perturbative series for inclusive @xmath0 decay rates in eq .
( [ schemrate ] ) ( or equivalently for the @xmath2 quark pole mass ) .
it was shown in ref .
@xcite that this cancellation of the linear ir sensitivity occurs similarly to all orders in @xmath17 .. it was argued in ref .
@xcite that this assumption may not hold beyond order @xmath134 . to order @xmath134 , i.e. to order @xmath39 ,
this assumption is proven by explicit calculations @xcite . ]
thus it is shown that using the upsilon expansion for @xmath0 decay rates eliminates all linear infrared sensitivity at least to order @xmath39 .
this is the order to which most @xmath0 decay rates have been computed , and the order to which we will present results in secs .
iv and v. while we can not prove that this cancellation is present at all orders in perturbation theory , this is likely in view of the fact that it holds both at large orders in the bubble summation approximation and in the first two orders including non - abelian contributions .
consequently , the upsilon expansion is insensitive to any order @xmath9 ambiguity in the @xmath2 quark mass which is present in perturbation theory .
an important theoretical uncertainty in applying the upsilon expansion is the size of nonperturbative corrections to the @xmath24 mass in eq .
( [ upsmass ] ) .
the dynamics of the @xmath24 system can be described using nrqcd @xcite .
the leading nonperturbative corrections to @xmath62 arise from matrix elements in the @xmath24 of @xmath135 , the hamiltonian of the light degrees of freedom . in @xmath0 mesons ,
the leading nonperturbative correction to the @xmath0 meson mass is due to the matrix element of @xmath135 , which is the @xmath136 parameter of order @xmath9 .
the @xmath9 dependence is different for the @xmath24 .
@xmath135 is the integral of a local hamiltonian density , @xmath137 the radius of the @xmath24 is @xmath138 , so the matrix element of @xmath135 is of order @xmath139 , by dimensional analysis .
( note that the matrix element of @xmath135 is of order @xmath140 , not @xmath141 .
terms that grow with @xmath34 can be treated using nrqcd perturbation theory . )
using @xmath142gev , and @xmath143mev , of order a constituent quark mass , gives a nonperturbative correction of @xmath144mev .
using instead @xmath145mev gives a correction of @xmath146mev .
we will use @xmath147mev as a conservative estimate of the nonperturbative contribution to @xmath62 . for large enough quark mass ,
the dominant nonperturbative corrections to the quarkonia masses are believed to come from the gluon condensate , @xmath148 .
its contribution to the mass of the @xmath149 @xmath150-wave state as calculated in refs .
@xcite is @xmath151 where @xmath17 in the denominator should be evaluated at the scale of the order of the bohr radius of the @xmath152 state , @xmath153 , and @xmath154 , @xmath155 , etc .
, are numerical constants .
values of the gluon condensate in the literature vary in the range @xmath156 , and so the uncertainties in eq .
( [ nonpert ] ) are sizable . in ref .
@xcite , eq . ( [ nonpert ] ) is estimated to give a 60 mev contribution to the @xmath1 mass .
equation ( [ nonpert ] ) seems to be enhanced by @xmath157 compared to our naive estimate of @xmath158 above .
this is due to the fact that @xmath159 and not @xmath160 is renormalization group invariant . in our estimate
the operator @xmath161 is renormalized at the scale @xmath162 .
the renormalization group scaling from @xmath162 to a low scale @xmath163 gev gives an additional factor @xmath164 , which is parameterically of order @xmath165 . since @xmath166 gev , we have not included this factor in our estimate .
our equations depend on the @xmath24 mass , and make no reference to the wave function of the @xmath24 .
in fact , there are much larger nonperturbative contributions to the @xmath1 wave function , which nevertheless , do not affect the energy of the state @xcite .
the reason is that the energy can be obtained in terms of the matrix element of the nrqcd effective hamiltonian , whereas the wave function has no simple operator expansion .
the situation is similar to other well - studied processes , such as the cross - section for @xmath167 , where the total cross - section has an ope description and much smaller nonperturbative corrections than other non - ope observables such as event shapes @xcite . from the results presented in ref .
@xcite , one can see three indications that the nonperturbative corrections to eq .
( [ upsmass ] ) may be smaller than typical expectations based on eq .
( [ nonpert ] ) .
first , the sign of @xmath168 is positive , and a large value for it increases @xmath14 as extracted from the inclusive semileptonic @xmath0 width .
thus , for a very large nonperturbative correction to @xmath62 ( say , above @xmath169mev or so ) , the consistency of the inclusive and exclusive @xmath14 determinations is significantly worse .
second , in ref .
@xcite we determined @xmath14 using both the mass of the @xmath1 and @xmath170 .
( [ nonpert ] ) shows that nonperturbative corrections to the @xmath152 @xmath150-wave state scale as @xmath171 .
if we assume that eq .
( [ nonpert ] ) holds at least approximately also for the size of nonperturbative effects for higher radial excitations of the @xmath24 , the nonperturbative contributions would be 64 times larger for the @xmath172 than for the @xmath78 .
the fact that the resulting values of @xmath14 coincide at the 5% level at order @xmath39 @xcite indicates that there can not be a large nonperturbative correction to the mass of the @xmath1 .
third , the `` most nonperturbative '' that an @xmath24 can be is for it to look like a @xmath173 bound state , which occurs just at threshold for decay into two mesons . the nonperturbative contribution to the @xmath174 mass is @xmath175 , and is of order @xmath176 gev . assuming again that this is approximately the nonperturbative contribution for @xmath177 , one finds that the @xmath1 state has a nonperturbative contribution of order @xmath178 mev , an absurdly small value .
finally , note that one can not separate perturbative and nonperturbative contributions to a physical quantity , since the perturbation series is a divergent series .
growth in the coefficients due to infared renormalons is removed , but there are other sources of factorial growth in the coefficients , such as the number of diagrams at @xmath152 order @xcite . ]
one can treat the perturbation series as an asymptotic series , and sum the terms until they no longer decrease in magnitude .
the resulting difference from the complete answer ( which is not known ) can be called the nonperturbative contribution . if the perturbation series is poorly behaved , so that one can keep only a few terms in the asymptotic expansion
, then one would expect that the error in the asymptotic expansion is large
, that nonperturbative effects are also large . to make this more precise , consider the @xmath80 total cross - section into hadrons at @xmath179 , in the limit of massless quarks . by dimensional analysis , the cross - section must have the form @xmath180 the dimensionless function @xmath181 includes both perturbative and non - perturbative effects , and has an asymptotic series expansion in @xmath17 .
for an asymptotic expansion , the best approximation to @xmath182 is given by summing the series till the terms start to increase .
the error at the minimum term , which is typically of the form @xmath183 can be considered the nonperturbative contribution to @xmath182 .
the terms in the @xmath17 expansion for the @xmath24 mass in eq.([upsmass ] ) start increasing already at order @xmath39 . using the order @xmath49 term as an estimate of the error in the asymptotic series gives @xmath184 mev , which can be taken as an estimate for the size of nonperturbative contributions .
when one combines @xmath62 with @xmath0 decay rates , the perturbation series is much better behaved , and one expects a much smaller nonperturbative contribution to @xmath0 decay rates expressed in terms of @xmath62 . in other words , the dominant part of the large nonperturbative contributions to @xmath62
are closely related to the @xmath9 uncertainty in the pole mass .
what is relevant for our analysis is the nonperturbative contributions to @xmath62 after these pole mass effects have been removed .
the residual contributions are expected to be much smaller , of order @xmath185 , and formally of the same order as @xmath186 corrections which have been neglected in the inclusive decay rates .
it is these residual nonperturbative effects for which we have used the estimate of 100 mev . to summarize
, we think that the size of nonperturbative corrections to @xmath62 relevant for our analysis are hardly known at the order of magnitude level ( the conventional estimates yield hundreds of mev , but our arguments above indicate that they could equally be tens of mev only ) .
the best way of determining them is directly from the experimental data .
we would like to find @xmath0 decay observables sensitive to @xmath62 for which parameter - free predictions can be made .
the best candidates are shape variables in inclusive decays or form factor ratios in exclusive decays .
the reason is that observables like total inclusive rates are only moderately sensitive to @xmath62 , and a possible discrepancy between experiment and theory could always be interpreted as a different value for a ckm angle , as opposed to constraining nonperturbative corrections to @xmath62 .
the most promising observables are the ones which are the most sensitive to @xmath136 : the mean photon energy in @xmath187 , and the form factor ratio @xmath188 in exclusive @xmath189 decay .
these are discussed in secs .
v.b and iv , respectively .
exclusive semileptonic @xmath189 decays depend on four form factors , @xmath190 where @xmath191 . in the infinite mass limit @xmath192 and the other three are equal to the isgur - wise function @xcite .
one linear combination of the form factors is not measurable when the lepton masses are neglected .
it is conventional @xcite to define two ratios of the three measurable form factors @xmath193 here we shall concentrate on the prediction for @xmath194 at order @xmath195 in the upsilon expansion . using the notation and results of ref .
@xcite , the theoretical prediction for @xmath196 is @xmath197 + \ldots \ , , \qquad\ ] ] where the elipses denote the known order @xmath198 and unknown order @xmath199 corrections .
if the qcd sum rule prediction @xmath200 @xcite is approximately correct , then the @xmath201 correction is much smaller than the @xmath202 term .
one can avoid relying on a theoretical estimate for @xmath203 , since the deviation of @xmath204 from unity measures @xmath205 directly if the @xmath206 subleading isgur - wise function is neglected .
since @xmath206 parameterizes @xmath207 corrections due to insertion of the chromomagnetic operator in the lagrangian , it is expected to be small and a qcd sum rule calculation supports this @xcite . neglecting @xmath206 seems to be less model dependent than relying on the previously mentioned result for @xmath203 .
the relation between @xmath208 and @xmath203 when @xmath206 is neglected is @xmath209\,{2m_c \over \bar\lambda ( 1 + 3z)},\ ] ] where @xmath210 . the perturbative correction @xmath211 is known to order @xmath212 @xcite , @xmath213 + \bigg ( { \bar\alpha_s \over 4\pi } \bigg)^2 \beta_0 c_f \bigg [ \frac56\ , \phi(z ) - \frac73 \bigg ] \,,\ ] ] where @xmath214 and @xmath215 from results given in ref .
@xcite , we can obtain the coefficient @xmath216 to order @xmath217
@xmath218 + \bigg ( { \bar\alpha_s \over 4\pi } \bigg)^2 \beta_0 c_f \bigg [ \frac{13}6\ , \phi(z ) + \frac{25}3 \bigg ] \,.\ ] ] to combine these results with eq .
( [ upsmass ] ) , it seems simplest to reexpress @xmath219 in terms of @xmath61 .
this yields @xmath220 thus we obtain at order @xmath195 in the upsilon expansion @xmath221 \,.\ ] ] the quantity @xmath12 appears in this result because we used eq .
( [ mbmc ] ) to eliminate the dependence on the charm quark mass .
the last term in square brackets is the @xmath201 contribution in eq .
( [ r1exp ] ) , and we have assumed that it is small , so that terms of order @xmath49 or @xmath12 from this contribution have been omitted . a @xmath147mev nonperturbative contribution to @xmath62 changes the prediction for @xmath196 by @xmath222 , which is about twice the order @xmath195 correction .
probably the largest uncertainty is due to the neglected @xmath199 corrections @xcite , which need to be better understood before eq .
( [ r1pred ] ) may be trusted at the percent level .
one can compare eq .
( [ r1pred ] ) with the conventional treatment .
the size of the order @xmath217 terms is usually quoted in terms of the blm scale .
the blm scale of @xmath211 is @xmath223 and that of @xmath224 is @xmath225 @xcite .
surprisingly , the blm scale of @xmath216 is much smaller , @xmath226 , and that of @xmath227 in eq .
( [ cratio ] ) is only slightly larger , @xmath228 .
these low blm scales show that the conventional perturbation series for @xmath196 is poorly behaved .
the perturbation series in eq .
( [ r1pred ] ) can not be converted into a blm scale since it mixes different orders in @xmath17 . however , to compare with conventional results one can define an effective blm scale by using the ratio of the @xmath195 and @xmath49 terms , just as the conventional treatment used the ratio of the @xmath229 and @xmath17 terms . the effective blm scale of eq .
( [ r1pred ] ) is @xmath230 , which is large and allows for a reliable expansion .
the present experimental results are @xcite @xmath231 this is consistent with eq .
( [ r1pred ] ) , although the errors are sizable . if @xmath232 will be measured in the future with one ( or at most a few ) percent errors , then it may be possible to constrain nonperturbative contributions to @xmath62 at the @xmath147mev level , which would be very interesting . to summarize , eq .
( [ r1pred ] ) predicts the form factor ratio @xmath196 independent of @xmath136 and the quark masses .
if this is tested and works at the few percent level , then we can make better and still model independent predictions for @xmath0 decay form factors in general , including the ones to the orbitally excited mesons @xmath233 and @xmath234 @xcite .
even if nonperturbative contributions to @xmath62 can not be constrained too well , our results show that expressing @xmath0 decay rates in terms of @xmath62 is a useful and theoretically consistent parameterization .
in this section , we apply the upsilon expansion to inclusive @xmath0 decays .
the @xmath2 quark decay rate has long been known for arbitrary masses of the four fermions participating in the weak decay , both at tree level @xcite and at order @xmath17 @xcite .
the order @xmath235 nonperturbative corrections will be taken into account for all decay modes , however , @xmath17 corrections to these will be neglected .
the numerical results below correspond to the choice @xmath58 and @xmath60 , and @xmath37-dependence will be discussed as well .
nonleptonic decays are more complicated than semileptonic decays .
it is less clear whether local duality holds in nonleptonic decays , since there is no smearing variable @xcite .
there are also corrections not present in semileptonic decay because the four - quark operators run in the effective theory below the weak scale .
while the complete series of leading @xcite and subleading @xcite logarithms have been summed , we can not consistently use these results , since that would involve summing series of the form @xmath236^n$ ] and @xmath237^n$ ] ( where @xmath238 ) without summing @xmath239 ( where @xmath240 ) .
instead , we follow an approach somewhat similar to ref .
@xcite based on the observation that the largest corrections arise at low orders in @xmath17 .
this is quite reasonable , as the order @xmath241 ^ 2 $ ] term dominates the sum of leading logs ( there is no order @xmath242 term due to color conservation ) , and the order @xmath17 term dominates the sum of next - to - leading logs . for @xmath243 nonleptonic
decays we include the order @xmath17 correction , and at order @xmath109 the terms containing @xmath244 ^ 2 $ ] , @xmath245 , and @xmath38 .
terms might also be unexpectedly large due to coulomb enhancement @xcite . ] for @xmath246 nonleptonic decays we include the order @xmath17 and order @xmath241 ^ 2 $ ] corrections only .
implications of our method for semileptonic @xmath0 decays were studied in ref .
@xcite , and here we repeat the results briefly for completeness .
estimates of the small non - blm contributions to the decay rates at order @xmath109 , which were taken into account in ref .
@xcite , will be neglected here . for massless leptons ,
the @xmath17 correction to free quark decay is known analytically @xcite , and the order @xmath247 correction has been determined in ref .
@xcite .
the @xmath20 decay rate at order @xmath195 in the upsilon expansion is @xcite @xmath249 , \end{aligned}\ ] ] the perturbation series , @xmath250 , is far better behaved than the series in terms of the pole mass , @xmath251 , or the series expressed in terms of the @xmath22 mass at @xmath58 , @xmath252 . the uncertainty in the @xmath0 decay rate using eq .
( [ buups ] ) is much smaller than that in eq .
( [ schemrate ] ) , both because the perturbation series is better behaved , and because the @xmath24 mass is better known ( and better defined ) than the @xmath2 quark mass .
( [ buups ] ) yields a relation between @xmath15 and the total semileptonic @xmath20 decay rate with very small uncertainty , ) and ( [ vcb ] ) , we have included the non - blm terms of order @xmath39 as well ( see ref .
@xcite ) . ]
@xmath253 where we have used @xmath254 and @xmath255 .
the first error is obtained by assigning an uncertainty in eq .
( [ buups ] ) equal to the value of the @xmath39 term and the second is from assuming a @xmath147mev uncertainty in eq .
( [ upsmass ] ) . the scale dependence of @xmath15 due to varying @xmath37 in the range @xmath256 is less than 1% .
the uncertainty in @xmath12 makes a negligible contribution to the total error .
it is unlikely that @xmath257 will be measured without significant experimental cuts , for example , on the hadronic invariant mass @xcite .
our method should reduce the uncertainties in such analyses as well .
the @xmath19 decay rate at order @xmath195 is @xcite @xmath259 \ , , \end{aligned}\ ] ] for comparison , the perturbation series in this relation when written in terms of the pole mass is @xmath260 .
equation ( [ bcups ] ) implies @xmath261 where @xmath262 is the electromagnetic radiative correction .
the uncertainties come from assuming an error in eq .
( [ bcups ] ) equal to the @xmath39 term , the @xmath263 error in @xmath12 , and a @xmath147mev error in eq .
( [ upsmass ] ) , respectively .
the second uncertainty is reduced to @xmath264 by extracting @xmath12 from the electron spectrum in @xmath19 ; see eq .
( [ lambda1 ] ) .
the agreement of @xmath14 with other determinations ( such as exclusive decays ) is a check that nonperturbative corrections to eq .
( [ upsmass ] ) are indeed small . in ref .
@xcite @xmath136 and @xmath12 were extracted from the lepton spectrum in @xmath19 decay . with our approach
, there is no dependence on @xmath136 , so we can determine @xmath12 directly with small uncertainty . considering the observable @xmath265 , a fit to the same data yields @xmath266 the central value includes corrections of order @xmath212 @xcite .
the first error is dominated by @xmath186 corrections @xcite .
we varied the dimension - six matrix elements between @xmath267 , and combined their coefficients in quadrature in the error estimate .
the second error is from assuming a @xmath147mev uncertainty in eq .
( [ upsmass ] ) .
the central value of @xmath12 at tree level or at order @xmath17 is within @xmath268 of the one in eq .
( [ lambda1 ] ) .
this value of @xmath12 can be used consistently in theoretical expressions up to order @xmath247 @xcite .
the order @xmath235 corrections are taken from @xcite , and the order @xmath247 corrections from @xcite .
the @xmath272 decay rate at order @xmath195 is @xmath273 \ , , \end{aligned}\ ] ] for comparison , the perturbation series in this relation when written in terms of the pole mass is @xmath274 . for the @xmath275 decay rate we find @xmath276 \,,\ ] ] for comparison ,
the order @xmath49 correction written in terms of the pole mass is @xmath277 .
the ratio @xmath278 is interesting because it directly probes lepton mass effects , and it can be predicted precisely in the standard model .
the largest uncertainty in the prediction is @xmath12 , and varying it in the range given in eq .
( [ lambda1 ] ) yields @xmath279 . , and neglected its uncertainty .
] thus , in the standard model , measurements of @xmath271 can be used to constrain @xmath12 @xcite .
the most recent aleph data , @xmath280 @xcite , implies at the @xmath281 level @xmath282 .
however , at the @xmath283 level this upper bound already becomes positive , which is not very interesting since @xmath12 is expected to be negative .
@xmath271 is also interesting for constraining new physics , e.g. , possible ( tree level ) charged higgs contributions in the large @xmath284 region @xcite .
the mean photon energy over the region @xmath286 in inclusive @xmath187 decay for sufficiently small @xmath287 is sensitive to @xmath136 and independent of @xmath288 when the decay rate is expressed in terms of @xmath289 and @xmath136 instead of @xmath34 @xcite .
we expect that stringent constraints can be obtained on the nonperturbative corrections to @xmath62 when a measurement of the photon energy spectrum is available with @xmath287 lowered to around @xmath290gev @xcite .
our method reduces the theoretical uncertainties in the total @xmath187 decay rate only moderately .
the @xmath187 decay rate is usually normalized to the semileptonic @xmath19 rate to eliminate the @xmath23 factor from the prediction .
the two biggest ingredients of the total theoretical error are the scale dependence of the @xmath19 rate computation itself ( without the @xmath23 factor ) , and to a lesser extent the normalization to the @xmath19 rate due to the uncertainty of @xmath291 and the experimental error .
the order @xmath235 corrections are taken from @xcite .
we neglect the strange quark mass , and the deviation of @xmath292 and @xmath293 from unity .
the most interesting decay mode is @xmath294 .
, , \end{aligned}\ ] ] the variable @xmath296 is introduced to distinguish between terms that come from different powers of @xmath245 .
equation ( [ bccsd ] ) has been written with the renormalization group scaling factor expanded in powers of @xmath297 .
one can effectively include the complete leading logarithmic series by replacing the @xmath298 term by @xmath299 .
this replacement also holds for eqs .
( [ bcusd ] ) , ( [ buusd ] ) and ( [ bucsd ] ) . the upsilon expansion does not affect the order @xmath300 and @xmath301 terms , i.e. , they are the same as in the conventional @xmath17 expansion .
the behavior of the order @xmath49 and @xmath195 terms have improved significantly . when written in terms of the pole mass , these terms are @xmath302 @xcite .
the cancellation we find is rather nontrivial .
the expansion of @xmath23 when reexpressed in terms of @xmath62 generates positive contributions at order @xmath303 .
this is why there was a significant cancellation in the perturbation series for @xmath304 decay , since the conventional perturbation series expansion for this decay had negative @xmath303 terms . on the other hand , the conventional perturbation series for nonleptonic @xmath305 has positive @xmath303 terms , and reexpressing the overall @xmath23 factor in terms of @xmath62 only makes them larger .
it is the expansions of the tree level @xmath294 phase space , which generates large negative contributions at order @xmath303 to make the expansion reasonably well - behaved . for nonleptonic @xmath306 decay ,
the rate is @xmath307 \ , , \end{aligned}\ ] ] when written in terms of the pole mass , the order @xmath49 and @xmath195 terms are @xmath308 .
the decays mediated by @xmath246 transitions are much less important , and therefore we keep fewer terms than before . @xmath309 \,,\ ] ] for comparison , the order @xmath49 correction written in terms of the pole mass is @xmath310 .
@xmath311 \,,\ ] ] this is the only case where the upsilon expansion increases the order @xmath49 correction compared to when it is written in terms of the pole mass , @xmath312 . the theoretical prediction for the semileptonic @xmath0 branching ratio @xmath313 was discussed in @xcite .
it has been emphasized repeatedly that the calculation of the @xmath294 rate is the least reliable . with our expansion in eq .
( [ bccsd ] ) , the ratio of the @xmath314 correction to the order @xmath49 term is significantly smaller than that calculated in terms of the pole masses , so the @xmath294 rate should be under better control . with our central values for the inclusive decay rates given in the previous section we obtain @xmath315 for the semileptonic branching ratio and the average charm multiplicty , which is consistent with other recent analyses . here
we have used @xmath316 .
the sensitivity of these results to @xmath12 , for example , is small when it is varied in the range obtained in eq .
( [ lambda1 ] ) .
if we had not expanded the series of leading logarithms , @xmath317 would remain unchanged , but @xmath318 would increase to @xmath319 . including a @xmath320 mev nonperturbative contribution to the @xmath24 mass
changes these values by a negligible amount .
we have shown that inclusive and exclusive @xmath0 decay rates can be predicted in terms of the @xmath1 mass instead of the @xmath2 quark mass .
it is crucial to our analysis to use the upsilon expansion in @xmath49 rather than the conventional expansion in powers of @xmath17 .
our formulae relate only physical quantities to one another .
they result in smaller theoretical uncertainties than existing numerical predictions , and the behavior of the perturbation series is improved .
moreover , the uncertainties can be estimated without resorting to cumbersome arguments , and they can be checked using the experimental data . in a previous paper @xcite , we had shown how our method works for inclusive semileptonic @xmath0 decays . in this article , we have shown how it applies to exclusive decays and to nonleptonic decays . in the new cases studied , the upsilon expansion also improves the behavior of the perturbation series .
the biggest theoretical uncertainty at present is a reliable bound on the effect of nonperturbative contributions to the @xmath24 mass ( other than the linearly infrared sensitive piece of the pole mass ) . the best way of determining this seems to be directly from the experimental data .
the comparison of @xmath321 extracted using the upsilon expansion with previous determinations already suggests that these contributions are not large .
we are investigating whether other processes can be used to provide an independent test of this result .
we thank mike luke and especially mark wise for useful discussions , and for trying to keep us honest .
this work was supported in part by the department of energy under grant doe - fg03 - 97er40546 , and by the national science foundation under grant phy-9457911 .
fermilab is operated by universities research association , inc .
, under doe contract de - ac02 - 76ch03000 .
g. altarelli and s. petrarca , phys .
b261 ( 1991 ) 303 ; + i. bigi , b. blok , m.a .
shifman , and a. vainshtein , phys .
b323 ( 1994 ) 408 ; + e. bagan , p. ball , v.m .
braun , and p. gosdzinsky , phys .
b342 ( 1995 ) 362 [ ( e ) b374 ( 1996 ) 363 ] ; + m. neubert and c.t .
sachrajda , nucl .
b483 ( 1997 ) 339 . | theoretical predictions for @xmath0 decay rates are rewritten in terms of the @xmath1 meson mass instead of the @xmath2 quark mass , using a modified perturbation expansion .
the theoretical consistency of this expansion is shown both at low and high orders .
our method improves the behavior of the perturbation series for semileptonic and nonleptonic inclusive decay modes , as well as for exclusive decay form factors .
the results are applied to the determination of the semileptonic @xmath0 branching ratio , charm counting , the ratio of @xmath3 and @xmath4 decay rates , and form factor ratios in @xmath5 decay .
we also comment on why it is not possible to separate perturbative and nonperturbative effects in qcd . |
density functional theory ( dft ) @xcite has become a powerful tool for physicists and chemists to describe the electronic structure of atoms , molecules , and solids . while being exact in theory , practical applications of dft rely on approximations of the exchange - correlation ( xc ) functional . the local spin density approximation ( lsd ) @xcite is in most cases too crude an approximation for the study of molecular systems .
functionals based on the generalized gradient approximation ( gga ) @xcite allow for a more accurate description of the inhomogeneous electron densities of molecules but also turn out not to be accurate enough in many cases .
the accuracy of gga functionals for applications of chemical interest can be further improved by admixture of a fraction of exact exchange in the form of hybrid functionals @xcite .
the b3lyp @xcite hybrid functional has become widely used in molecular studies . more recently
a number of new functionals have been developed to reproduce certain chemical or molecular properties to high accuracy , while being less accurate for others @xcite . in the spin - unrestricted kohn - sham ( ks ) formalism @xcite the energy of a system of @xmath0 electrons in an external potential @xmath1
is defined through the spin - densities @xmath2 by @xmath3 = t_{\text{s}}[{\rho^{\uparrow}},{\rho^{\downarrow } } ] + v_{\text{ext}}[\rho ] + e_{\text{h}}[\rho ] + e_{\text{xc}}[{\rho^{\uparrow}},{\rho^{\downarrow}}]\ , \ ] ] where @xmath4 is the electrostatic interaction energy of the density with the external field and @xmath5 is the hartree energy , the classical coulomb repulsion of the charge density with itself .
@xmath6 is the kinetic energy of the non - interacting reference system , constructed from the set of ks - orbitals @xmath7 to produce the same density as the exact wave function .
the exchange - correlation energy , @xmath8 , contains all remaining contributions to the exact energy .
the xc - energy can be split into the sum of exchange , @xmath9 , and correlation energy , @xmath10 . for any one - electron density , @xmath11 ,
the two conditions @xmath12 = - e_{\text{h}}[\rho^{1 } ] \label{eq : xc_cond_x}\ ] ] and @xmath13 = 0 \label{eq : xc_cond_c}\ ] ] are fulfilled by the exact functional .
the second condition can be satisfied by a semi - local form of the correlation functional as the one proposed by lee , yang , and parr ( lyp ) @xcite .
it is , however , not possible to formulate an exchange functional that can , for any possible @xmath11 , compensate the non - local hartree energy merely from local information of the density . for approximate semi - local functionals condition ( [ eq : xc_cond_x ] ) , or both conditions , ( [ eq : xc_cond_x ] ) and ( [ eq : xc_cond_c ] ) , are violated .
this gives rise to a self - interaction error ( sie ) , @xmath14 = e_{\text{h}}[\rho^{1 } ] + e_{\text{xc}}[\rho^{1},0]\ .
\label{eq : sie}\ ] ] perdew and zunger proposed a self - interaction correction scheme ( sic ) in which the sie of the individual orbitals , as defined by eq .
( [ eq : sie ] ) , is subtracted from the total energy @xcite .
the perdew - zunger self - interaction correction ( pz - sic ) energy functional , @xmath15 = e^{\text{ks}}[{\rho^{\uparrow}},{\rho^{\downarrow } } ] - \sum_{i=1}^{n } e^{\text{sie}}[\rho_{i}]\ , \ ] ] depends not only on the total spin - densities , but also on the orbital densities , @xmath16 with @xmath17 , and can in principle be applied to any approximate xc - functional . for the exact functional , the correction term vanishes for any many - electron system , as can be seen from eqs .
( [ eq : xc_cond_x ] ) and ( [ eq : xc_cond_c ] ) . for approximate functionals ,
the pz - sic is accurate for any one - electron density but for many - electron systems it is in general only an approximate correction , as the magnitude of the many - electron self - interaction error does not have to be the sum of the individual sie terms of eq .
( [ eq : sie ] ) @xcite .
while this orbital based estimate can not be expected to eliminate all self - interaction for many - electron systems , it may improve the accuracy of approximate functionals .
the purpose of the present study is to test the accuracy of pz - sic for a few commonly used functionals when applied to molecules .
errors in the energy due to approximation of the xc - functional stem from both an incomplete cancellation of the electron self - interaction and an inaccurate description of the inter - electronic interaction .
the accuracy of the pz - sic does , therefore , depend on the functional approximation it is used with . from earlier studies ,
it was concluded that pz - sic often overcorrects errors in calculated observables of many - electron systems such as equilibrium bond lengths and atomization energy .
a scaled down modification of the pz - sic functional has been proposed in the form of @xmath18 = e^{\text{ks}}[{\rho^{\uparrow}},{\rho^{\downarrow } } ] - \alpha \sum_{i=1}^{n } e^{\text{sie}}[\rho_{i}]\ , \label{eq : scaled_sic}\ ] ] and for the perdew , burke , and ernzerhof ( pbe ) functional a factor @xmath19 in the range of 0.4 to 0.5 could improve the description of some observables @xcite .
more elaborate scaling schemes have also been proposed @xcite .
a recent study of atoms with pz - sic showed that the total energy can be lowered significantly by allowing the orbitals to be complex functions @xcite .
this improved significantly the results when pz - sic was applied to the pbe functional , while a restriction to real orbitals led to much larger errors than those of the uncorrected functional@xcite . according to the variational principle
, the addition of an imaginary component to the orbitals can only lower the total energy .
the effect on equilibrium geometry or energy differences such as atomization energy or energy barriers is not monotonous in a similar way . to our knowledge ,
previously published fully variational studies of the energetics of molecules using stationary pz - sic have exclusively been based on real orbitals .
only recently complex orbitals were used in a study focusing on bond - lengths of molecules within an approximate kohn - sham interpretation of the pz - sic @xcite .
we present results on the ground state geometry and atomization energy of a set of 17 molecules , the equilibrium structure of two ` problematic ' radicals and the energy barrier of four reactions .
we studied the effect of sic using three different functionals : the local spin density approximation ( using slater exchange @xcite and the perdew - wang parameterization of correlation @xcite , spw92 ) , the two generalized gradient approximations of perdew , burke , and ernzerhof @xcite , and becke s exchange functional @xcite and correlation of lee , yang , and parr @xcite ( blyp ) .
the results of self - consistent calculations using pz - sic ( sic ) as well as the scaled down modification with a factor of one half ( sic/2 ) are presented . for comparison ,
calculations using the two hybrid functionals , pbe0 @xcite and b3lyp @xcite , were also carried out , as well as less accurate sic calculations using real orbitals .
the energy minimum with respect to variation of the orbitals under the constraint of orthonormality is described by the two sets of equations @xcite , @xmath20 where the orbital - specific hamiltonians are defined by @xmath21 the lagrange multipliers can be determined by projection of eq .
( [ eq : sic_min_cond ] ) as @xmath22 .
in contrast to semi - local functionals , the matrix of lagrange multipliers is not hermitian for any unitary transformation among the occupied orbitals .
the minimum energy is determined both by the space spanned by the set of orbitals and by the linear combination of the orbitals .
the effective potential will not be the same for all orbitals and this places sic outside the domain of kohn - sham dft .
it can be treated as a true kohn - sham functional by means of the optimized effective potential method ( oep ) @xcite , but in many applications the functional is treated in the generalized kohn - sham framework , i.e. , the energy is minimized with respect to variation of the orbitals , resulting in different potentials for each one of them .
analytical gradients of the energy of sic functionals have been derived @xcite and can be used in a direct minimization of the energy .
the efficiency can be improved by an additional step in the iterative minimization . before the orbitals are altered according to the energy gradient , the unitary transformation that minimizes the sic energy terms is found .
by such a ` unitary optimization ' the convergence rate can be greatly improved @xcite .
an efficient algorithm for the unitary optimization has recently been developed @xcite .
the calculations were carried out with the gaussian - type orbital based program quantice @xcite . as analytical gradients of the atomic positions are not available in the program , the molecular structure has been optimized manually .
a sequence of internuclear distances and angles was sampled and the minimum energy configuration found by cubic interpolation .
the equilibrium structure was confirmed by comparison of the interpolated energy with the calculated energy at the interpolated geometry .
this manual scheme limits the size of the molecules and the number of structural degrees of freedom that can be optimized in the current version of this software . in the geometry optimization ,
all analogous bonds in a molecule were constrained to have the same length to reduce the computational effort compared to a completely unconstrained structural relaxation . for all calculations , atom - centered grids of 75 radial shells @xcite of a 302-point lebedev - laikov grid @xcite
were combined to form a multicenter integration grid @xcite .
the cartesian representation of correlation - consistent polarized valence - triple - zeta @xcite ( cc - pvtz ) basis sets were used for most calculations . for the reaction barriers involving only hydrogen ,
a quadruple - zeta basis set @xcite ( cc - pvqz ) was used .
the convergence criterion of the electronic minimization was @xmath23 for the norm of the energy gradient .
all calculations were using unrestricted orbitals , starting from a random initialization .
the orbital density dependence of the pz - sic energy expression results in a set of well defined ` optimal orbitals , ' defined by a unitary transformation of the canonical orbitals . for atomic systems ,
these orbitals have often been found to resemble @xmath24 hybrid orbitals @xcite . when the orbitals are allowed to be complex functions ,
the optimal orbitals still resemble hybrid orbitals but can have significantly different shape and orientation than hybrids of real orbitals . for neon ,
e.g. , real orbitals produce a set of @xmath25 orbitals with tetrahedral orientation , but complex orbitals produce a set with tetragonal orientation without nodal surfaces @xcite . for molecules ,
the optimal orbitals often take forms that are consistent with ` chemical intuition , ' i.e. they can be interpreted as lone pairs , single , or multiple bonds . also
, orbitals consistent with the more ` exotic ' three - center or banana bonds can form . also here , the shape of the complex orbitals is often rather different from that of the real ones , but the interpretation above can still be retained in many cases .
figure [ fig : n2_orbitals ] shows the optimized valence orbitals of one spin channel for , obtained from pbe+sic using complex ( a ) and real ( b ) orbitals .
the real orbitals are of two kinds : three spatially degenerate orbitals that add up to a triple bond , and two lone pairs .
the triple bond is built up from two different kinds of orbitals .
one orbital has the character of a sigma - bond with rotational symmetry about the molecular axis .
the other two orbitals are degenerate and symmetric about a plane going through the molecular axis
. the complex lone pairs do not differ much qualitatively from the real orbitals .
the real orbitals are rather localized and in staggered orientation , but the complex orbitals are more delocalized and to a larger extent share the same space .
the ground state energy and equilibrium geometry of the molecules , , , , hf , , , , , co , no , , , , , , and triplet methylene , , was calculated . to test the accuracy of the various density functional approximations , the predicted atomization energy , i.e. , the difference in the total energy of the atoms constituting a molecule and the total energy of the molecule , was calculated and compared with experimental estimates corrected for zero - point energy @xcite or , in the case of h@xmath26 , with an accurate theoretical result@xcite . the mean error ( me ) and mean absolute error ( mae ) for each functional approximation is shown in figure [ fig : bond_energy ] . for comparison ,
results of sic and sic/2 calculations restricted to real orbitals are also shown .
the numerical values are listed in table [ tab : bond_energy ] , except for the results obtained with real orbitals .
the lsd energy shows the largest deviation of all functionals , strongly over binding the molecules .
the errors are reduced by applying sic ( see lsd+sic ) , but this only partially eliminates the errors .
the gga functionals , pbe and blyp , reduce the errors of lsd significantly , but still predict most molecules to be too stable .
the binding energy is reduced by sic for both functionals , but the correction is too large .
for pbe+sic , the mae is actually slightly increased and is doubled for blyp+sic . the mean deviation
is greatly reduced by applying sic scaled by one - half , pbe+sic/2 , while the mae is just slightly smaller than for the uncorrected functional .
blyp+sic/2 gives smaller errors than blyp+sic but still predicts the binding energy to be too low .
calculations using sic that are restricted to real orbitals predict lower atomization energy on average .
vydrov _ et al .
_ studied the effect of pz - sic on the heat of formation using several functionals using calculations restricted to real orbitals @xcite .
they concluded that pz - sic only improves the results for lsd , while larger deviations are found when the correction is used with gga functionals .
as shown here , it is important to allow the orbitals to be complex functions .
this reduces the over correction , but does not eliminate it .
better agreement with the reference data is obtained by scaling the sic .
no fitting of the scaling factor was carried out , but a factor of one - half chosen to illustrate the trend .
the scaled sic used with the pbe functional gives a smaller mean error than the pbe0 hybrid functional , however , the mae indicates that some molecules are over bound while others are too unstable . a systematic under binding
is found for molecules containing hydrogen , except for and .
lrrrrrrrrrrrr @xmath27 ( ev ) & & & & & & & & & & & & + & 4.75 & 0.16 & 0.18 & 0.22 & -0.21 & -0.26 & -0.30 & -0.22 & 0.00 & -0.04 & -0.07 & 0.03 + & 1.13 & -0.10 & -0.10 & -0.09 & -0.27 & -0.25 & -0.24 & -0.30 & -0.24 & -0.25 & -0.25 & -0.23 + & 2.52 & 0.11 & 0.13 & 0.17 & -0.21 & -0.24 & -0.26 & -0.24 & -0.01 & -0.07 & -0.12 & 0.01 + & 6.03 & 0.73 & 0.41 & 0.14 & -0.01 & -0.43 & -0.78 & -0.34 & 0.06 & -0.41 & -0.80 & -0.11 + & 6.16 & 0.79 & 0.50 & 0.25 & -0.08 & -0.38 & -0.63 & -0.29 & -0.13 & -0.45 & -0.72 & -0.20 + & 9.84 & 1.75 & 1.38 & 1.05 & 0.73 & 0.33 & 0.00 & -0.07 & 0.59 & 0.02 & -0.46 & 0.11 + & 5.12 & 2.49 & 1.51 & 0.65 & 1.15 & 0.27 & -0.45 & 0.30 & 0.81 & -0.22 & -1.06 & 0.27 + & 1.65 & 1.78 & 0.70 & -0.15 & 0.71 & -0.60 & -1.52 & -0.08 & 0.55 & -1.02 & -1.65 & 0.01 + & 5.03 & 1.14 & 1.00 & 0.89 & 0.18 & 0.16 & 0.18 & -0.27 & 0.19 & 0.07 & 0.01 & -0.08 + & 11.32 & 1.66 & 1.16 & 0.70 & 0.37 & -0.04 & -0.39 & -0.23 & 0.07 & -0.47 & -0.92 & -0.23 + & 6.63 & 2.00 & 1.50 & 1.06 & 0.85 & 0.34 & -0.02 & 0.02 & 0.61 & -0.04 & -0.50 & 0.10 + & 17.00 & 3.55 & 2.36 & 1.29 & 1.12 & 0.21 & -0.54 & 0.01 & 0.42 & -0.70 & -1.59 & -0.15 + & 18.21 & 1.84 & 1.91 & 2.04 & 0.03 & -0.23 & -0.43 & -0.09 & -0.10 & -0.46 & -0.74 & 0.06 + & 12.88 & 1.67 & 1.47 & 1.33 & 0.16 & -0.34 & -0.74 & -0.14 & 0.14 & -0.41 & -0.84 & 0.11 + & 10.10 & 1.35 & 0.94 & 0.58 & -0.03 & -0.51 & -0.92 & -0.34 & -0.11 & -0.59 & -0.99 & -0.18 + & 8.20 & 1.01 & 1.06 & 1.12 & 0.24 & 0.18 & 0.16 & 0.20 & 0.04 & -0.03 & -0.08 & 0.13 + & 17.52 & 2.41 & 2.34 & 2.29 & 0.49 & 0.40 & 0.35 & 0.02 & 0.07 & -0.06 & -0.13 & -0.03 + me & & 1.43 & 1.09 & 0.80 & 0.31 & -0.08 & -0.38 & -0.12 & 0.17 & -0.30 & -0.64 & -0.02 + mae & & 1.44 & 1.10 & 0.82 & 0.40 & 0.30 & 0.46 & 0.19 & 0.24 & 0.31 & 0.64 & 0.12 + [ tab : bond_energy ] an extreme case of the under binding obtained from the blyp+sic functional is , which is predicted to be unstable .
also , pbe+sic gives severe underestimation of the binding energy and elongation of the bond as shown in table [ tab : f2_bond ] .
the bond energy is overestimated by all uncorrected functionals . for lsd ,
sic/2 and sic reduce the bond energy , the latter giving a value closer to experiment . for pbe , the correction greatly reduces the binding energy , resulting in an underestimation already for sic/2 and predicting a very weakly bound molecule for sic .
this effect is even more pronounced for blyp , where sic predicts the molecule not to be bound at all .
usually , sic/2 gives results that are intermediate of those obtained by sic and by the uncorrected functional .
this is not the case , however , for in the gga functionals .
the bond length is reduced by sic/2 , but increased when full sic is applied .
the optimal orbitals obtained using pbe+sic/2 and pbe+sic are qualitatively different , as shown in figure [ fig : f2_bond ] .
the two orbitals ( one each for spin - up and spin - down ) corresponding to the single bond are similar for pbe+sic/2 .
the slight difference in shape does not result in spin polarization of the total density , as it is compensated by the density of the lone pairs .
for pbe+sic , the orbitals are distorted and localize one on each of the two nuclei .
the total density is spin polarized and the electronic structure can be interpreted as an intermediate state towards two separated fluorine atoms .
the effect of sic on the molecule and the single atoms is not balanced .
the total energy of the molecule is predicted to be too high relative to that of the atoms , resulting in a weak bond .
for blyp+sic the ` correction ' is unbalanced to such an extent that the energy of is above that of the atoms for all nuclear separations .
building the bond in a diatomic molecule can be described as a delocalization of atomic hybrid orbitals over both nuclei , accompanied by changes in the shape of the orbitals to maintain orthogonality as well as relaxation of the orbitals not participating in the bond .
the sic energy of an orbital varies strongly depending on how localized it is .
the valence atomic orbitals become more compact with increasing atomic number within a row of the periodic table . at the same time , the magnitude by which sic reduces the atomization energy , increases for all functionals going from to to . moving from the second to the third row of the periodic table ,
the valence orbitals become more delocalized , and the changes in atomization energy due to sic are smaller as can be seen by comparing with .
preliminary results of the binding energy of larger molecules at unrelaxed geometry reveal similar trends .
these trends indicate that the effect of sic on atomization energy is more pronounced if localized atomic orbitals participate in the bonding . however , without taking into account the changes in the orbital shape and the rearrangement of non - bonding orbitals , such a simplified interpretation is insufficient to explain all the observed effects and a more detailed study is required .
lrrrclrrr ( ev ) & & & & & ( ) & & & + ref .
& 1.65 & 1.65 & 1.65 & & ref .
& 1.41 & 1.41 & 1.41 + uncorr . &
3.43 & 2.36 & 2.20 & & uncorr . &
1.38 & 1.41 & 1.43 + sic/2 & 2.35 & 1.05 & 0.63 & & sic/2 & 1.33 & 1.36 & 1.38 + sic & 1.50 & 0.13 & 0.00 & & sic & 1.30 & 1.44 &
+ hybrid & & 1.57 & 1.66 & & hybrid & & 1.38 & 1.40 + [ tab : f2_bond ] calculations of observables from energy differences usually are more accurate than the individual total energy values because of partial cancellation of errors .
one source of errors is the limited basis set used in the calculations .
this error can in theory be eliminated by systematically increasing the size of the basis set until a complete basis is reached , or in a more practical way , until the calculated energy differences do not change significantly .
even if a complete basis set is used , an error remains from the approximate energy functional .
a functional with large errors in the predicted total energy must be seen as too crude an approximation of the exact functional .
this , however , does not mean that it can not be a useful functional for practical calculations .
if the functional has a rather constant error per electron for all systems , the energy difference of systems of the same number of electrons still can be predicted accurately .
figure [ fig : total_energy ] correlates the errors in total energy of the molecules to that of the constituent atoms .
each circle depicts one molecule , the x - axis shows the error per electron of the molecule , the y - axis that of the atoms .
the diagonal dashed line corresponds to a perfect cancellation of errors , for points above or below the line , atomization energy is overestimated or underestimated , respectively .
the diameter of the circles indicates the number of electrons in each system .
the molecules containing hydrogen are indicated by circles filled in grey . in this representation
, the effect of sic and the admixture of exact exchange can be studied in more detail than from the atomization energy alone .
the total energy of both the atoms and the molecules is overestimated by pbe , around 0.2 ev per electron for most molecules .
the largest deviation is found for , the largest system of the test set . with the exception of ( the smallest point ) , the errors of the molecules and atoms are similar , but some spread around the diagonal is observed .
half sic and full sic reduce the absolute errors for most systems with most points moving closer to the origin .
however , the magnitude of the correction is in many cases different for atoms than molecules .
going from pbe to pbe+sic/2 to pbe+sic , the points are shifted down and the vertical spread is reduced , which corresponds to smaller and more similar errors per electron for the atoms . at the same time ,
however , the horizontal spread increases , indicating a less systematic correction of the errors in the total energy of the molecules . this can be observed in particular for the molecules containing hydrogen , indicated as grey points .
when pbe is used , the errors for the molecules are more systematic than the errors for the atoms .
the opposite trend is observed when pbe+sic is used . for pbe0 ,
the absolute magnitude of the errors becomes less systematic for both atoms and molecules , the spread is increased along the diagonal . at the same time , the spread perpendicular to the diagonal is reduced , improving the overall cancellation of errors . for blyp the total errors are much smaller than for pbe , and the cancellation is slightly better . blyp+sic/2 reduces the total errors for many systems , but underestimates the atomization energy . for sic ,
the points move further below the diagonal but also spread both vertically and horizontally , indicating an unsystematic effect of sic on the total energy of both atoms and molecules .
b3lyp actually increases the magnitude of the errors over blyp , predicting too low energy for all the atoms and molecules .
the errors are , however , well balanced and cancellation of errors results in the superior performance of b3lyp with respect to atomization energy , as shown in figure [ fig : bond_energy ] .
this improved cancellation of errors can to some extent be explained by the origin of this hybrid functional .
b3lyp is based on the b3pw91 functional , for which the parameters had been fit to reference data composed of the total energy of ten atoms and 106 energy differences @xcite .
such a fitting procedure places less weight on the accuracy of total energy than on cancellation of errors .
the same parameters are used in the b3lyp functional that , despite its different functional components , predicts energy differences often very accurately while atomic total energy is predicted to be too low .
[ cols= " < , > , > , > , > , > , > , > , > , > , > , > , > " , ] [ tab : barriers ] the energy barrier for four reactions has been calculated to study the effect of sic on the activation energy .
the energy difference between the reactant minimum and the lowest saddle point of the potential energy surface is calculated .
the reactions are listed in table [ tab : barriers ] as well as the geometry of the saddle point and the barrier height evaluated from _ ab initio _ calculations . for and
, the saddle point had been calculated using configuration interaction ( ci ) @xcite .
the barrier heights are calculated with respect to the reactants , using the accurate total energy of the hydrogen molecule @xcite and the exact energy of the hydrogen atom as reference values . for hfh , only the collinear ,
symmetric saddle point was considered , as a reference for this barrier is available @xcite , and the computational effort was too large to find the non - collinear saddle point which is slightly lower in energy @xcite .
the inversion barrier height of had been calculated using the coupled cluster method , ccsd(t ) @xcite . .
the deviation from reference energy , @xmath28 in ev , and bond lengths , @xmath29 in , is shown for the various functionals .
the rectangle , 1 pm by 0.1 ev , at the origin emphasizes the different energy and length scales of the different graphs . for hfh and
the results based on real orbitals are depicted by grey symbols , for and real and complex orbitals give identical results . ]
figure [ fig : barriers ] shows the deviation of the calculated saddle points from the reference values . for and ,
the barrier height predicted by the uncorrected functionals is always too low , and is increased by sic/2 and sic . for both reactions , sic improves the energy barrier but it is still underestimated for .
the bond lengths at the saddle point decrease for sic/2 and further for sic , as is also found for most molecules . compared to the bond lengths predicted by the uncorrected functionals , this increases the deviation for and overcorrects the slightly too large bonds for .
the barrier is underestimated by the uncorrected functionals and increases with sic , monotonously from sic/2 to sic , the latter giving an overcorrection . for the gga
functionals the bond lengths do not change monotonously but behave similar to the bond length discussed above .
the bond length decreases for sic/2 and agrees better with the reference value , but then increases for pbe+sic , while hardly changing for blyp+sic . for the functionals studied here , the gga+sic/2 barriers give the best results , being more accurate than the hybrid functionals . for this reaction , in contrast to the hydrogen barriers ,
different results are obtained when using real orbitals . here
, the bond length is larger and the barrier height is slightly lower than for complex orbitals .
these three barriers describe bond breaking situations and are underestimated by the uncorrected functionals .
the ammonia inversion is qualitatively different , as bonds merely rearrange and their lengths at equilibrium and saddle point do not differ significantly .
the calculated bond lengths in the planar configuration are elongated by less than 2 pm compared to the equilibrium structure @xcite .
the inversion barrier calculated with the uncorrected functionals is in good agreement with the reference , but the bonds are predicted to be too long .
the bond length decreases and the barrier increases with sic for lsd , but for the gga functionals the barrier is lowered and becomes underestimated . using real orbitals , on the other hand , the barrier increases strongly .
this qualitatively different effect of sic can be explained by the structure of the ` lone pair , ' as was done for the methyl radical . for the planar molecule ,
the orbitals of both spins look very similar to those in figure [ fig : ch3_orbitals ] . for the complex orbitals , spin up and spin down orbitals
are rotated with respect to each other by 120@xmath30 . in the real case ,
the spin - up and spin - down orbitals are symmetric with respect to reflection through the plane . as for , this configuration of the real orbitals is higher in energy than that of the complex orbitals , which explains the large difference in the predicted energy barrier . as for the atomization energy , the errors in the calculated reaction barriers
can be analyzed in greater detail by comparison of the errors in total energy of the species involved .
figure [ fig : total_energy_saddle ] shows the error per electron of the total energy of the saddle point structure ( * s * ) , the separated reactants ( * r * ) and that of the separated atoms ( * a * ) , calculated using pbe , pbe+sic/2 , and pbe+sic .
points on the dashed line correspond to a perfect description of the total energy , the vertical difference between neighboring points corresponds to deviations in the atomization energy ( a - r ) and energy barrier ( r - s ) . for the systems ,
the pbe energy of the hydrogen atoms is quite accurate and the saddle point energy is only slightly underestimated .
the underestimation of both the barrier height and atomization energy originates mainly from an inaccurate description of the hydrogen molecule .
when sic is applied , the energy of the atom is lowered insignificantly and the energy of the molecule is raised slightly , making both total and atomization energy of the molecule less accurate .
however , pbe+sic introduces a large error at the saddle point , which cancels the error in the molecule . for sic/2 ,
the errors introduced are less well balanced and the barrier is still underestimated .
the barrier is severely underestimated by the pbe functional stemming from equally large errors of opposite sign in the reactants and saddle point structure .
the energy of the saddle point is predicted accurately by pbe+sic .
each optimized orbital is to great extent localized at one of the hydrogen atoms and neighboring orbitals are of opposite spin . in such a configuration ,
most of the exchange - correlation energy can be expected to be self - interaction energy , in which case pz - sic usually performs well . in this case
the barrier is improved by sic , but as the energy of the hydrogen molecule is too high , a still significant error in the barrier remains . for the atomic reference of hfh and , both the error in the total energy and the change when applying sic are dominated by the heavy atoms .
the reduction of the error by sic is smaller for the hf molecule than for the atoms , thus making the binding energy less accurate compared to pbe .
the error in the saddle point energy is increased and exceeds that of the reactants . here ,
half sic gives a better cancellation of error than both pbe and pbe+sic .
for the molecules , the effect is smaller and the barrier height changes only slightly . for the reaction
, we have calculated the energy along a path connecting the saddle point and the reactants .
for a number of distances @xmath31 , the bond length of the fragment was relaxed . in figure
[ fig : h3_relaxed ] , the relaxed energy of the lsd and pbe functionals is shown for several separations of the fragments .
the leftmost points correspond to the bond length and energy barrier of the saddle point .
increasing values of @xmath31 describe the separation into a hydrogen atom and a molecule . in lsd ,
the combined system has lower energy than the separated fragments for all separations , the energy barrier is negative and is predicted to be a stable molecule . lsd+sic/2 and lsd+sic increase the energy and predict a reaction barrier .
the height is however still underestimated and at intermediate distances , a system more stable than the reactants is found . in pbe ,
no intermediate configuration is more stable than the reactants , but the energy difference is globally underestimated .
pbe+sic agrees well with the ci results while the pbe0 hybrid does not describe the system as well .
however , as shown in figure [ fig : total_energy_saddle ] the improved description by the pbe+sic functional results only from a better cancellation of errors and not from a better description of the system .
the results presented here on the energetics of small molecules provide insight into both the strengths and the shortcomings of the perdew - zunger self - interaction correction .
the qualitatively wrong equilibrium structure of predicted by pbe is corrected by sic . in the case of the radical
, sic had been found to predict an incorrect geometry , but in the present study this was shown to be an artifact of a restriction to real orbitals .
this molecule and the potential energy surface emphasize the importance of using sic with gga functionals to obtain more accurate results , as had already been shown in a study of the total energy of atoms @xcite .
however , not all gga functionals are suited for the application of sic .
the molecule is found to be unstable when sic is applied to the blyp functional and the errors in total energy are large and unsystematic for the rest of the molecules in the test set as well as the isolated atoms .
analysis of the total energy revealed that for pbe+sic , the errors per electron in the atomic systems are reduced and show less fluctuations than that of the molecules , resulting in an unsystematic effect on the atomization energy .
the perdew - zunger sic does not in general result in an overcorrection when applied to pbe , as the predicted total energy is still too high .
however , an overcorrection in calculated atomization energy is usually observed which can be improved by scaling the sic . the simple scaling scheme of using a constant factor of one - half for all orbitals when applied with pbe reduces the mean error in atomization energy to less than that of the pbe0 hybrid functional , but still gives significant absolute errors . for systems with only a few electrons
, this half - sic approximation does not perform better than full sic in calculations of energy barriers .
more flexible scaling @xcite where full sic is retained for one - electron systems or isolated orbitals could work better for such systems .
these scaling schemes have so far only been used in combination with real orbitals and their performances would need to be reassessed using complex orbitals .
this work was supported by the graduate fellowship fund of the university of iceland , the icelandic research fund , the marie curie research training network ( mcrtn ) `` hydrogen , '' and the nordic energy research network `` solar fuel . ''
the calculations were carried out on the computer clusters `` sol , '' funded by the icelandic research fund , and `` gardar , '' operated by the nordic high performance computing ( nhpc ) project . | self - consistent calculations using the perdew - zunger self - interaction correction ( pz - sic ) to local density and gradient dependent energy functionals are presented for the binding energy and equilibrium geometry of small molecules as well as energy barriers of reactions .
the effect of the correction is to reduce binding energy and bond lengths and increase activation energy barriers when bond breaking is involved .
the accuracy of the corrected functionals varies strongly , the correction to the binding energy being too weak for the local density approximation but too strong for the gradient dependent functionals considered . for the perdew , burke , and ernzerhof ( pbe ) functional , a scaling of the pz - sic by one half gives improved results on average for both binding energy and bond lengths .
the pz - sic does not necessarily give more accurate total energy , but it can result in a better cancellation of errors .
an essential aspect of these calculations is the use of complex orbitals .
a restriction to real orbitals leads to less accurate results as was recently shown for atoms [ s. klpfel , p. klpfel , and h. jnsson , phys .
rev .
a 84 , 050501 ( 2011 ) ] .
the molecular geometry of radicals can be strongly affected by pz - sic .
an incorrect , non - linear structure of the radical predicted by pbe is corrected by pz - sic .
the radical is correctly predicted to be planar when complex orbitals are used , while it is non - planar when the pz - sic calculation is restricted to real orbitals . |
one source of inspiration in symbolic dynamics comes from storage systems and transmission in computer science . for example
sofic shifts are analogous to regular languages in automata theory , so a sofic shift and its cover are natural models for information storage and transmission . as a result , starting with a presentation of a dynamical system , there are known algorithms constructed to compute some kind of object from such presentation . given a sofic shift , coven and paul constructed a finite procedure to obtain a finite - to - one sofic cover @xcite .
there is an algorithm to determine whether two graphs present the same sofic shift @xcite .
kim and roush showed that the shift equivalence of sofic systems is decidable @xcite . in this work ,
starting from a sofic shift and its finite - to - one cover , we present an algorithm to compute the number of preimages of a typical point of the sofic . moreover , we show that in the case of having an infinite - to - one cover , an analogous object can be computed in finitely many steps . given a factor code @xmath0 from a one - dimensional shift of finite type @xmath1 to a sofic shift @xmath2 , when @xmath0 is finite - to - one
there is a quantity assigned to @xmath0 called the _ degree _ of @xmath0 .
the degree of a finite - to - one code is defined to be the minimal number of @xmath0-preimages of the points in @xmath2 .
one can show that the number of preimages of every transitive point in @xmath2 is exactly the degree of @xmath0 .
the degree of a finite - to - one code is widely - studied and known to be invariant under recoding @xcite . in the first section of this paper we present an algorithm to compute this invariant . when @xmath3 is not limited to be finite - to - one an analogous of the degree , called the _ class degree _ ,
is defined to be the minimal number of transition classes ( always finite ) over the points in @xmath2 .
the definition of a transition class is motivated by communicating classes in markov chains .
roughly speaking , two preimages @xmath4 and @xmath5 of a point @xmath6 in @xmath2 lie in the same equivalence class , _ transition class _ , if one can find a preimage @xmath7 of @xmath6 which is equal to @xmath4 up to an arbitrarily large given positive coordinate and right asymptotic to @xmath5 and vice versa .
when @xmath0 is finite - to - one then the degree and the class degree of @xmath0 match .
one can also show that the class degree is invariant under topological conjugacy and the number of transition classes over any transitive point of @xmath2 is exactly the class degree of @xmath0 .
one of the main applications of the class degree is bounding the number of measures of relative maximal entropy @xcite .
such measures have applications in information theory , computing hausdorff dimensions and functions of markov chains @xcite . in the second section of this paper
we show that the class degree is computable .
throughout this paper , @xmath1 is a one - dimensional shift of finite type ( sft ) with the shift transformation @xmath8 .
the alphabet of @xmath1 is denoted by @xmath9 and the @xmath10-algebra on @xmath1 generated by cylinder sets is denoted by @xmath11 .
a triple @xmath12 is called a * factor triple * when @xmath3 is a continuous shift - commuting map ( factor code ) from an sft @xmath1 onto a subshift @xmath2 ( sofic shift @xmath2 ) .
when @xmath0 is a one - block factor code induced by a symbol - to - symbol map @xmath13 we naturally extend @xmath14 to blocks in @xmath11 ( b stands for block ) . when @xmath0 is a finite - to - one factor code there is a uniform upper bound on the number of pre - images of points in @xmath2 .
the minimal number of pre - images of points in @xmath2 is called the * degree * of the code and is denoted by @xmath15 .
we say two factor triples @xmath12 and @xmath16 are * conjugate * if @xmath1 is conjugate to @xmath17 under a conjugacy @xmath18 , @xmath2 is conjugate to @xmath19 under a conjugacy @xmath20 , and @xmath21 .
@xcite[thm : magicsymbol ] let @xmath12 be a factor triple .
there is a factor triple @xmath22 conjugate to @xmath12 such that @xmath23 is one - step and @xmath24 is one - block .
@xcite[invdeg ] given two conjugate factor triples @xmath12 and @xmath22 , we have @xmath25 .
[ thm : samenumber ] @xcite let @xmath0 be a finite - to - one factor code from an sft @xmath1 onto an irreducible sofic shift @xmath2 .
then every transitive point of @xmath2 has exactly @xmath26 preimages .
given a one - block factor code @xmath0 , above every @xmath2-block @xmath27 there is a set of @xmath1-blocks @xmath28 which are sent to @xmath27 by @xmath14 ; i.e. , @xmath29 . given @xmath30 , define @xmath31 and @xmath32 [ thm : degmag]@xcite let @xmath0 be a finite - to - one one - block factor code from an sft @xmath1 onto an irreducible sofic shift @xmath2 .
then @xmath33 given a one - block factor code @xmath3 , a * magic block * is a block @xmath27 such that @xmath34 for some @xmath35 .
such an index @xmath36 is called a * magic coordinate * of @xmath27 .
a factor code @xmath0 has a * magic symbol * if there is a magic block of @xmath0 of length @xmath37 .
the class degree defined below is a quantity analogous to the degree which is defined in the general case when @xmath0 is not only limited to be finite - to - one .
[ defn : equiv ] suppose @xmath12 is a factor triple and @xmath38 .
there is a * transition * from @xmath4 to @xmath39 denoted by @xmath40 if for each integer @xmath41 , there is a point @xmath42 in @xmath1 so that 1 .
@xmath43 , and 2 .
@xmath44 for some @xmath45 .
write @xmath46 , and say @xmath4 and @xmath39 are in the same ( equivalence ) * transition class * if @xmath47 and @xmath48 .
the minimal number of transition classes over points of @xmath2 is called the * class degree * of @xmath0 and denoted by @xmath49 .
[ invclass ] @xcite given two conjugate factor triples @xmath12 and @xmath22 , we have @xmath50 .
[ thm : classdegree ] @xcite let @xmath0 be a one - block factor code from a one - step sft @xmath1 to a sofic shift @xmath2 .
the number of transition classes over a right transitive point of @xmath6 is exactly the class degree .
theorem [ thm : classdegree ] , in below , provides a finitary characterization of the class degree .
[ defn : tb ] let @xmath51 be a one - block factor code from a one - step sft @xmath1 to a sofic shift @xmath2 and let @xmath52 be a block of @xmath2 .
let @xmath53 and let @xmath54 be a subset of @xmath55 .
we say @xmath56 is * routable * through @xmath57 at time @xmath41 if there is a block @xmath58 with @xmath59 , and @xmath60 .
a triple @xmath61 is called a * transition block * of @xmath0 if every block in @xmath62 is routable through a symbol of @xmath54 at time @xmath41 .
the cardinality of the set @xmath54 is called the * depth * of the transition block @xmath61 .
let @xmath63 a * minimal transition block * of @xmath0 is a transition block of depth @xmath64 .
[ thm : classdegree ] @xcite let @xmath0 be a one - block factor code from a one - step sft @xmath1 to a sofic shift @xmath2 .
then @xmath65 .
the following theorem shows the relation between the degree and the class degree of a finite - to - one factor code .
@xcite let @xmath51 be a finite - to - one factor code from a sft @xmath1 to an irreducible sofic shift @xmath2
. then @xmath66 .
in this section we present an algorithm to compute the degree of a finite - to - one factor code . by theorems [ thm : magicsymbol ] and
[ invdeg ] , without loss of generality , we may assume @xmath0 is a one - block factor code defined on a one - step sft .
let @xmath1 be a one - step sft with alphabet @xmath67 and adjacency matrix @xmath68 .
let @xmath3 be a finite - to - one one - block factor code from @xmath1 to a sofic shift @xmath2 with the alphabet @xmath69 .
make two graphs @xmath70 and @xmath71 as follows .
let @xmath70 and @xmath71 both have the same vertex set @xmath72 where @xmath73 stands for the power set of @xmath74 .
for what we need later , divide the vertex set @xmath75 into two parts @xmath76 and @xmath77 where @xmath78 .
form the adjacency matrix @xmath54 of @xmath70 and the adjacency matrix @xmath79 of @xmath71 as follows .
let @xmath80 , then @xmath81 and @xmath82 for some @xmath83 in @xmath84 .
say @xmath85 if the following conditions hold . 1 .
@xmath86 is a block of @xmath2 , 2 .
@xmath87 is exactly the set of all symbols @xmath88 in @xmath89 such that @xmath90 is a block of @xmath1 for some @xmath91 in @xmath92 .
otherwise @xmath93 .
say @xmath94 if we have 1 .
@xmath86 is a block of @xmath2 , 2 .
@xmath92 is exactly the set of all symbols @xmath91 in @xmath95 such that @xmath90 is a block of @xmath1 for some @xmath88 in @xmath87 .
otherwise @xmath96 .
consider the following subsets of the vertex set @xmath75 , @xmath97 and @xmath98 [ thm : degalg ] using above notations , we have @xmath99 note that @xmath70 and @xmath71 are finite directed graphs .
let @xmath100 and @xmath101 be the shift spaces represented by @xmath70 and @xmath71 accordingly .
let @xmath102 be the map from @xmath75 to @xmath2 taking @xmath103 to @xmath104 for some @xmath105 ( note that @xmath106 is independent of @xmath105 ) .
then @xmath102 induces a one - block code @xmath107 and a one - block code @xmath108 .
+ the key feature of these two graphs is the following .
for any block @xmath109 of @xmath2 there is a unique walk @xmath110 in @xmath70 with the following properties .
1 . @xmath111 .
2 . @xmath112 .
3 . @xmath113 .
similarly there is a unique walk @xmath114 in @xmath71 with the following properties .
1 . @xmath115 .
2 . @xmath116 .
3 . @xmath117 .
let @xmath118 , @xmath119 , and @xmath120 .
let @xmath121 be a walk in @xmath70 where @xmath122 and @xmath123 .
let @xmath124 be a walk in @xmath71 where @xmath125 and @xmath126 .
since @xmath120 we have @xmath127 , and consequently @xmath128 is a block of @xmath2 .
by the key feature discussed in the above paragraph , we have @xmath129 since by definition @xmath130 we have @xmath131 now we prove the inequality in the other direction ; that is we seek to show that @xmath132 let @xmath133 be a magic block of @xmath0 with a magic coordinate @xmath134 .
let @xmath135 consider @xmath136 and @xmath137 .
let @xmath138 be the unique path in @xmath70 which maps to @xmath28 and @xmath139 . then @xmath140 let @xmath141 be a unique path in @xmath71 which maps to @xmath142 and @xmath143
. then @xmath144 it follows that @xmath145 therefore @xmath146 which completes the proof .
in this section we find an upper bound on the length of a minimal transition block of a factor triple .
then by theorem [ thm : classdegree ] it follows that the class degree of a factor code is computable . without loss of generality , by theorem [ thm : magicsymbol ] and theorem [ invclass ]
we may assume that the sft is one - step and the factor code is one - block .
[ thm : length ] let @xmath3 be a one - block factor code from a one - step sft @xmath1 to a sofic shift @xmath2 .
let @xmath147 .
there is a minimal transition block @xmath61 of @xmath0 with @xmath148 we need the following definitions and lemmas to prove theorem [ thm : length ] .
[ defn : pairs ] let @xmath3 be a one - block factor code from a one - step sft @xmath1 to a sofic shift @xmath2
. given @xmath149 and @xmath150 say @xmath92 * pairs with * @xmath151 * in form * @xmath152 when there is a block @xmath153 of @xmath2 with @xmath154 , @xmath155 such that @xmath156 if and only if there is @xmath157 which begins at @xmath158 and ends at @xmath159 ; that is , @xmath160 and @xmath161 . note that @xmath92 can pair with @xmath151 in at most @xmath162 distinct forms .
[ defn : w - pair ] let @xmath3 be a one - block factor code from a one - step sft @xmath1 to a sofic shift @xmath2 .
given @xmath149 and a block @xmath153 of @xmath2 with @xmath154 , @xmath155 , a @xmath27-*paring of * @xmath163 , denoted by @xmath164 , is the set @xmath150 which contains all pairs @xmath165 such that there is a block in @xmath157 with @xmath160 and @xmath161 .
[ cl : join ] let @xmath3 be a one - block factor code from a one - step sft @xmath1 to a sofic shift @xmath2 .
let @xmath166 .
let @xmath167 and @xmath168 be two blocks of @xmath2 with @xmath169 , @xmath170 , and @xmath171 .
let @xmath172 and @xmath173 be two blocks of @xmath2 with @xmath174 , @xmath175 , and @xmath176 .
then we have @xmath177 where @xmath27 and @xmath178 are blocks of @xmath2 formed by joining @xmath179 to @xmath180 and @xmath181 to @xmath182 as follows : @xmath183 and @xmath184 . first note that @xmath185 and @xmath186 and therefore @xmath27 and @xmath178 are legal blocks of @xmath2 with @xmath187 and @xmath188 .
it is enough to show @xmath189 .
let @xmath190 be in @xmath191 .
this means there is a block @xmath68 in @xmath192 which starts at @xmath158 and ends at @xmath193 .
since @xmath194 we have @xmath195 .
denote @xmath196 by @xmath159 .
it follows that @xmath197 and @xmath198 .
hence by assumption , @xmath199 and @xmath200 .
this means there is a block @xmath201 with @xmath202 and @xmath203 , and a block @xmath204 with @xmath205 and @xmath206 .
form the block @xmath207 by joining the blocks @xmath142 to @xmath28 as below ; @xmath208 . clearly
@xmath209 . since @xmath210 and @xmath211 it follows that @xmath212 .
[ cl : transblock ] let @xmath61 be a transition block with @xmath213 .
let @xmath154 , @xmath155 , and @xmath214 for some @xmath166 .
let @xmath178 with @xmath215 be another block of @xmath2 with @xmath154 , @xmath216 for some @xmath217 , and @xmath218 such that @xmath219 and @xmath220 .
then @xmath221 is a transition block of @xmath0 .
note that by lemma [ cl : join ] , since @xmath219 and @xmath220 , we have @xmath222 .
suppose @xmath223 .
let @xmath224 , we need to show that @xmath28 is routable through a member of @xmath54 .
let @xmath202 and @xmath225 .
since @xmath212 we have @xmath226 .
it follows that there is a block @xmath227 with @xmath228 and @xmath229 .
since @xmath61 is a transition block , then block @xmath142 must be routable through some symbol @xmath230 which implies that @xmath231 , and @xmath232 .
it follows that there is a block @xmath233 with @xmath234 , @xmath235 , and @xmath236 , meaning that @xmath28 is routable through @xmath230 .
let @xmath61 be a minimal transition block of @xmath0 with @xmath213 .
assume @xmath237 .
we construct a transition block @xmath238 with @xmath239 , but yet @xmath240 .
such transition block has depth @xmath241 and therefore is a minimal transition block .
let @xmath154 , @xmath155 .
denote @xmath242 by @xmath152 .
we show that @xmath92 can pair with @xmath151 in form @xmath152 in less than or equal @xmath243 numbers of steps .
recall that given any @xmath244 there are at most @xmath245 number of distinct subsets of @xmath246 . moreover , given @xmath247
, @xmath92 can pair with @xmath180 in at most @xmath248 distinct forms .
therefore , since @xmath249 , there is at least one symbol @xmath244 and a subset @xmath247 such that @xmath250 occurs in @xmath27 at two distinct positions @xmath251 and @xmath252 , with @xmath253 , and @xmath254 .
let @xmath255 be the block @xmath256 .
note that by assumption , @xmath257 which implies that @xmath258 is a legal block of @xmath2 .
if @xmath259 then we are done , if not , we repeat the process until @xmath92 pairs with @xmath180 in form @xmath152 in less than @xmath243 number of steps .
the result is a block @xmath260 of @xmath2 where @xmath261 , @xmath262 , @xmath263 , and @xmath264 .
now let @xmath214 , and denote @xmath265 by @xmath266 .
by similar argument @xmath151 can pair with @xmath267 in less than or equal @xmath243 number of steps ; that is , there is a block @xmath268 of @xmath2 where @xmath269 , @xmath270 , @xmath271 , and @xmath272 .
make the new block @xmath273 by joining the two blocks @xmath70 and @xmath274 as follows : @xmath275 .
note that since @xmath276 the block @xmath178 is a legal block of @xmath2 .
moreover , we have @xmath277 .
now it is easy to see that @xmath178 is a minimal transition block .
note that @xmath278 , @xmath279 , @xmath280 , @xmath281 , and @xmath282 .
lemma [ cl : transblock ] implies that @xmath221 is a transition block of @xmath0 , and since its depth is @xmath241 we conclude that it is a minimal transition block of @xmath0 .
given a factor triple @xmath12 since @xmath283 , if @xmath284 then @xmath285 .
thus the algorithm given in section [ sec : degalg ] will compute the class degree as well .
we can actually use the same two graphs @xmath70 and @xmath71 and modify theorem [ thm : degalg ] by applying the upper bound on the length of a minimal transition block stated in theorem [ thm : length ] to write an algorithm to compute the class degree .
however , since the bound on theorem [ thm : length ] is growing exponentially with respect to the number of preimages of a symbol in @xmath84 , the algorithm could be hopelessly complicated .
it would be ideal to find an efficient algorithm for computing the class degree .
d. blackwell .
the entropy of functions of finite - state markov chains . in _ transactions of the first prague conference on information theory , statistical decision functions , random processes held at liblice near prague from november 28 to 30 , 1956
_ , pages 1320
. publishing house of the czechoslovak academy of sciences , prague , 1957 .
k. petersen .
information compression and retention in dynamical processes . in _ dynamics and randomness ( santiago , 2000 )
_ , volume 7 of _ nonlinear phenom .
complex systems _ , pages 147217 .
kluwer acad .
publ . , dordrecht , 2002 . | let @xmath0 be a factor code from a one dimensional shift of finite type @xmath1 onto an irreducible sofic shift @xmath2 .
if @xmath0 is finite - to - one then the number of preimages of a typical point in @xmath2 is an invariant called the degree of @xmath0 . in this paper we present an algorithm to compute this invariant .
the generalized notion of the degree when @xmath0 is not limited to finite - to - one factor codes , is called the class degree of @xmath0 . the class degree of a code is defined to be the number of transition classes over a typical point of @xmath2 and is invariant under topological conjugacy .
we show that the class degree is computable . |
the interest in the studies of clusters is largely because of their possible technological applications which include the possibilities of developing novel cluster - based materials using the size dependence of their properties .
doping of clusters is an important possibility in this direction . in recent times
the fabrication of alloy clusters of different sizes with well defined , controlled properties by varying the composition and atomic ordering , has caught considerable attention.@xcite bimetallic alloy clusters have been known and exploited for last few years in various applications , specially in the catalytic reactions.@xcite varying the ratio of the two constituents , one can alter the surface structures , compositions and segregation properties.@xcite in this way , it is possible to tune the chemical reactivity at the surface of an alloyed cluster.@xcite considering the huge possibility of using nano clusters in catalysis , the study of cluster reactivity has become an interdisciplinary topic of present day research.@xcite about two decades ago , nonose _
et al _ measured the reactivity towards h@xmath3 of bimetallic co@xmath4v@xmath2 ( @xmath5 @xmath6 @xmath7 ) clusters using laser vaporization technique and reported strong cluster size and composition dependence.@xcite both v and co are 3@xmath8 transition metals .
the substitution of co by v atoms , one by one , should increase the reactivity of the alloy cluster towards h@xmath3 molecules , as v , an early transition metal , has a high reactivity towards h@xmath3 in contrast to co which has relatively low reactivity.@xcite interestingly , it was found that the reactivity increased gradually , as expected , with the substitution of co atom by v atom one by one in the co@xmath4 clusters ( @xmath5 @xmath9 13 ) , but for the co@xmath0 , there was a remarkable decrease in reactivity when a single co atom was substituted by a v atom . however , the reactivity increased as the number of exchange v - atoms increased further up to @xmath10 3 , while the fourth v - atom substitution did not increase the reactivity any more . in view of the high reactivity of elemental
v , this sudden drop in reactivity of the co@xmath11v cluster was rather surprising .
the authors speculated a plausible icosahedral structure for the co@xmath11v cluster with the active v atom at the cage center .
therefore , the v atom , being shielded geometrically from @xmath12 by the twelve surface co atoms , might have less chance to interact with h - atoms . the chemisorption reactivity of cationic co@xmath1v@xmath13 clusters @xcite and anionic co@xmath1v@xmath14 clusters @xcite also shows similar type of variation as that of neutral co@xmath1v@xmath2 clusters , which hints towards the dominant effect of geometric structure as compared to the electronic structure . for clusters ,
the ionization potential ( ip ) depends on the position of the highest occupied molecular orbital ( homo ) level . for the pure metal clusters , whetten _
et al _ have postulated that the reaction rate for cluster - h@xmath3 dissociative chemisorption is determined by the charge transfer from the homo to the lowest unoccupied molecular orbital ( lumo ) of the reactant gas h@xmath3 , and an anti - correlation between ip and reaction coefficient could be observed.@xcite that means lower value of ip corresponds to higher reactivity and vice versa . however , the ionization energies of the co@xmath4v@xmath2 clusters , measured by hoshino _ et al _ using photo - ionization spectroscopy show no such anti - correlation,@xcite again demonstrating the importance of geometrical structure .
a rigorous first - principles study in terms of geometric and electronic effects is therefore very much needed to understand this anomalous nature of reactivity of co@xmath1v@xmath2 ( @xmath10 0 - 4 ) clusters . in this report , we have carried out an _ ab - initio _ theoretical study on v doped co@xmath0 clusters .
the whole study can be divided into three major parts : first part consists of an exhaustive search for the minimum energy structures ( mes ) for cluster of each composition , followed by stability analysis of these mes in terms of various physical quantities in the second part , while the final part includes investigation of chemisorbed structures and understanding of cluster reactivity .
the calculations have been performed using density functional theory ( dft ) , within the pseudopotential plane wave method as implemented in vasp code.@xcite we have used the projector augmented wave ( paw ) pseudopotentials @xcite and the perdew - bruke - ernzerhof ( pbe ) exchange - correlation functional @xcite for spin - polarized generalized gradient approximation ( gga ) .
the 3@xmath8 and 4@xmath15 electrons were treated as the valence electrons for the transition metal elements and the wave functions were expanded in the plane wave basis set with the kinetic energy cut - off of 335 ev . the convergence of the cluster energies with respect to the cut - off value has been checked .
reciprocal space integrations were carried out at the @xmath16 point .
symmetry unrestricted geometry and spin optimizations were performed using the conjugate gradient and the quasi - newtonian methods until all the force components were less than a threshold value of 0.005 ev / . in order to ensure the structural trends found in the optimized structures , we have also carried out bohn - openheimer molecular dynamics @xcite within lda for few specific clusters .
simple cubic super - cells were used for cluster calculations with the periodic boundary conditions , where two neighboring clusters were kept separated by at least 12 vacuum space to make the interaction between the cluster images negligible .
the cohesive energy ( @xmath17 ) of a co@xmath4v@xmath2 alloy cluster is calculated with respect to the free atoms as @xmath18 where @xmath19(co@xmath4v@xmath2 ) , @xmath19(co ) and @xmath19(v ) are the total energies of co@xmath4v@xmath2 cluster , an isolated co atom and an isolated v atom respectively .
one can also define the cohesive energy with respect to the co and v bulk phases at equilibrium instead of isolated atoms .
however , the cohesive energy variation with increasing v - concentration follows the same trend irrespective of the way of cohesive energy definition .
the second difference in energy for fixed size ( @xmath20 constant ) and variable composition is defined as @xmath21 it gives the relative stability of alloy clusters having nearby compositions .
since chemical reaction with clusters takes place on its surface , the atomic arrangement and the composition of cluster surface play an important role in chemical reactivity .
therefore , the first step towards the theoretical modeling of clusters is to determine their ground state structure . in our earlier
work,@xcite we studied the structure and magnetism of the pure co clusters in detail . to obtain the mes having the optimized geometry as well as the optimized magnetic moment , we considered several probable starting geometries having closed packed atomic arrangement and allowed each of the geometries to relax for _ all possible _ collinear spin configurations of the atoms .
we follow the same way of structural optimization here for the v doped co clusters ( co@xmath1v@xmath2 ; @xmath10 1 - 4 ) . however , the situation for alloy clusters is quite cumbersome as one has to deal with a large number of starting geometries because of the presence of _ homotops _ ( the term was first coined by jellinek @xcite ) .
homotops _ have the same number of atoms , composition and geometrical structure , but differ in the arrangement of the doped atoms . as for example
, a single geometrical structure of an a@xmath4b@xmath2 alloy cluster with fixed number of atoms ( @xmath22 ) and composition ( @xmath23 ratio ) , will give , in principle , @xmath24 _ homotops_. some of them , however , may be symmetry equivalent and the number of the inequivalent _ homotops _ will be somewhat less than the above mentioned value .
so the variety of structures in alloy clusters is much richer than that of the pure clusters and the potential energy surface of even a small cluster of few tens of atoms is of enormous complexity .
it is generally found that the transition metal clusters prefer compact geometries to maximize the interaction between the rather localized @xmath8 orbitals.@xcite there are three most compact and highly coordinated structures for 13 atoms cluster : icosahedron , cub - octahedron and hexagonal closed packed ( hcp ) geometries .
we have therefore considered these three geometries as the most probable starting structures . for the pure co@xmath0 cluster
, we found that the mes is a distorted hcp geometry with total magnetic moment of 25 @xmath25 and total cohesive energy of 42.63 ev as mentioned in our previous work.@xcite the structure has 22 triangular faces and 33 edges ( cf .
[ fig : strco13 ] ) .
this hexagonal growth of the pure co clusters is quite unusual compared to the trend seen in the clusters of its 3@xmath8 neighbors like mn , fe , ni , which preferably adapt an icosahedral pattern.@xcite further study is needed to explore the unusual structural pattern of small pure co clusters.@xcite another distorted structure of hcp motif with total magnetic moment of 27 @xmath25 and 0.14 ev above the minimum energy state , is the first isomer .
the optimal icosahedral structure of total spin 31 @xmath25 ( structure having 20 triangular faces and 30 edges ) is 0.17 ev higher than the minimum energy state and emerges as the second isomer .
the third isomer is a distorted cub - octahedron with total magnetic moment 25 @xmath25 and it is 0.22 ev above the minimum energy state . the structures of the ground state , second and third isomers are shown in fig .
[ fig : strco13 ] .
it is to be noted that for the pure co@xmath0 cluster , all the atomic moments are ferromagnetically coupled and our predicted magnetic moment of the ground state structure is in good agreement with the experimental value which is 2.0 @xmath25/atom.@xcite in the case of single v atom doped co@xmath11v clusters , we have again considered the starting geometries of icosahedral , hcp and cub - octahedral symmetries .
however , depending upon the position of the singly doped v - atom either at the center or on the surface , it can give rise to several _
homotops_. again , the surface atoms of a 13-atoms cluster are not equivalent in terms of number of neighbors and their relative positions in case of hcp and cuboctahedral structures , while for icosahedral structure , all the surface atoms are equivalent
. this consideration will give few more starting geometries .
so we have considered all these possible geometrical structures and for each of them , we have considered the _ all possible _ collinear spin alignments during relaxation . upon optimization of geometry and spin degrees of freedom , we find that an icosahedral structure of total magnetic moment 25 @xmath25 , with v atom doped at the center position , is the mes . now there are several interesting points to note in the context of the co@xmath11v cluster .
first of all , the cohesive energy of the mes is considerably higher ( by 1.29 ev ) than that of the pure co@xmath0 clusters , indicating its exceptional stability .
secondly , the v doped co - clusters are found to prefer the icosahedral growth pattern , instead of hcp growth pattern , as we observed in case of the pure co - clusters .
therefore , just the single v - atom doping in co@xmath0 changes the equilibrium structure from hcp to icosahedron .
this hcp to icosahedron structural transition has also been reported previously for mn doping in co@xmath0 cluster.@xcite finally , the single v - atom likes to fit at the center of the icosahedron and it is ferrimagnetically coupled with the surface co - atoms . the first isomer is also a central v - atom doped icosahedral co@xmath11v cluster , with total magnetic moment 23 @xmath25 and it is about 0.46 ev above the mes .
center v doped cub - octahedral co@xmath11v cluster with 23 @xmath25 magnetic moment which has energy 0.70 ev higher than the minimum energy icosahedral structure , is the second isomer .
the third isomer is a center v doped icosahedron with total magnetic moment of 27 @xmath25 .
center v doping in case of hcp cluster is less favorable and appears as the fifth isomer in our calculation with total magnetic moment 25 @xmath25 and energy 0.80 ev above the mes .
we have also considered the optimal structure of co@xmath0 ( distorted hcp ) as the starting structure and substituted the most coordinated central co - atom by a v - atom .
after relaxation considering _ all possible _ spin alignments , it is found that the shape of the optimal structure remains more or less same as that of the optimal hcp co@xmath0 cluster and energetically , it is the fourth isomer with total magnetic moment 23 @xmath25 . the most probable surface v - doping has a cub - octahedral structure with total magnetic moment 21 @xmath25 , but it is about 1.32 ev above the mes .
the structure of ground state and few isomers for the co@xmath11v cluster are shown in fig .
[ fig : strco12v ] .
because of the icosahedral growth preference of the co@xmath11v cluster and the presence of the singly doped v - atom at the central site , we consider only the icosahedral symmetry based geometries as the starting guess for more than one v atom doped clusters , in which one v atom is always at the center position , while the residual v atoms reside on the surface . for co@xmath26v@xmath3 cluster
, the second v atom can replace any of the surface co atoms , as all the surface sites of a 13-atom icosahedron , are equivalent .
after relaxation for _ all possible _ spin configurations , we find that the mes has total magnetic moment of 19 @xmath25 and total cohesive energy of 43.22 ev .
the first and second isomers have magnetic moments of 21 @xmath25 and 17 @xmath25 and they are 0.10 ev and 0.25 ev above the minimum energy state respectively . in the mes ,
the center v - atom is again ferrimagnetically coupled with the surface co atoms and has much lower local magnetic moment as it was in the case of the co@xmath11v cluster , while the surface v atom has the maximum local magnetic moment and it is also ferrimagnetically coupled with the other surface co atoms . for co@xmath27v@xmath28 cluster , depending upon the different positions of the two surface v atoms , there are three possible icosahedral structures as shown in fig .
[ strco10v3 ] .
however , the structure of type @xmath29 where the two surface v atoms are closest to each other , appears as the most favorable one .
this type @xmath29 icosahedral structure with total magnetic moment of 21 @xmath25 is the minimum energy state with ferrimagnetic alignment of the central v - atom and ferromagnetic alignment of the surface v - atoms with the surface co - atoms .
both the central and the surface v - atoms have small values of local magnetic moments in this case .
there are several closely spaced isomers of type @xmath29 icosahedron with magnetic moments of 19 @xmath25 , 13 @xmath25 , 23 @xmath25 and 17 @xmath25 which are just 0.04 ev , 0.08 ev , 0.09 ev and 0.15 ev above the mes .
we find that the optimal structures of type @xmath30 and type @xmath31 are 0.20 ev and 0.31 ev above the mes , respectively and both of them have same total magnetic moment of 13 @xmath25 . for co@xmath32v@xmath33
, there are three v atoms on the surface . considering two surface v atoms at closest to each other ( following the ground state configuration of co@xmath27v@xmath28 ) , different positions of the third surface v atom
can give rise to the four icosahedral configurations as shown in fig .
[ strco9v4 ] . after optimization
, it is found that the type @xmath29 is the most favorable structure where all the three surface v atoms are closest to each other and form an octahedron with the central v atom .
the optimal structure of type @xmath29 icosahedron has total magnetic moment of 15 @xmath25 and total cohesive energy of 42.35 ev .
another type @xmath29 icosahedrons with magnetic moments of 13 @xmath25 , 19 @xmath25 , 21 @xmath25 and 17 @xmath25 are just 0.03 ev , 0.07 ev , 0.09 ev and 0.13 ev above the mes .
the optimal type @xmath30 , type @xmath31 and type @xmath34 icosahedrons have total magnetic moments of 15 @xmath25 , 13 @xmath25 and 15 @xmath25 respectively and they are 0.15 ev , 0.25 ev and 0.33 ev above the mes . the results of the structural optimizations , therefore , can be summarized as , unlike an hcp growth pattern of the pure co cluster , the v doped co@xmath0 clusters prefer to adopt an icosahedral packing . in such clusters ,
the most coordinated central site is occupied by a v - atom , while the residual v - atoms sit on the surface .
the surface v - atoms like to be closer to each other to form a group , thereby imparting more distortion to the structure and significantly alter the local surface charge density . the central v - site in the mes of all compositions , is always ferrimagnetically coupled and has lower magnetic moment . on the other hand
, the surface v - atoms can be ferrimagnetically or ferromagnetically coupled with the surface co - atoms and their local magnetic moments can be as low as the central v - site or as high as the surface co - site , depending upon the distribution of surface charge density in the clusters having different amount of v - doping .
our prediction of icosahedral geometry for the minimum energy states of v doped co clusters is in accordance with the speculation of a closed shell geometry around the central site in ref .
13 and ref . 15 .
+ to have more confidence about the structural trend as observed above , we have carried out few _ ab - initio _ molecular dynamics ( md ) study for the co@xmath11v and co@xmath27v@xmath28 clusters . to determine the lowest energy structures , we have heated up the clusters to 2000 k ( which is close to the melting temperature 1768 k for bulk co and 2183 k for bulk v ) .
the clusters have then been maintained at this temperature for at least 6 ps and then allowed to cool again to 0 k. the cooling process has been done for 24 ps and maintained a very slow cooling rate of 1 k per one iteration throughout the process .
the main results obtained from the md run are : ( a ) it prefers to adopt the icosahedral pattern with same magnetic alignment of atoms as we predicted from zero temperature relaxation , ( b ) for the co@xmath11v cluster , the v atom occupies the central position , ( c ) the surface v atoms for the co@xmath27v@xmath28 cluster , prefer to be closest to each other .
these results provide reassurance and more confidence in our zero temperature relaxation .
the observed atomic arrangement of the mes of the co@xmath1v@xmath2 clusters as described in the previous section depends critically on the balance of several parameters like the relative strengths of various kinds of bonds present in a cluster structure , the relative atomic sizes , the amount of charge transfer between two different species of atoms , the energy gap between the homo and the lumo , abundance of states near fermi energy etc .
below , we try to understand the structural stability of the clusters in terms of these parameters . fig .
[ fig : cohesive ] shows the plot of the cohesive energies of the mes of the co@xmath1v@xmath2 clusters with increasing v concentration .
the plot is with respect to the cohesive energy of the mes of co@xmath0 cluster .
it is seen that the binding has increased considerably compared to that of the optimal co@xmath0 cluster with single v atom doping .
however , the cohesive energies decrease sharply for the higher concentration of v - atoms . for the double and triple v - atom doping ,
the cohesive energies are above the dashed line indicating their higher stability compared to the pure co@xmath0 , while for the fourth v - atom doping , the cohesive energy is even lower than that of the pure co@xmath0 .
the relative stability among the clusters of nearby concentrations is more distinct when we plot the second difference in cohesive energies in the inset of fig .
[ fig : cohesive ] .
we use eqn .
[ d2e ] to calculate the second difference .
a sharp pick in the @xmath35 at @xmath10 1 points to the exceptional stability of the single v - doped cluster compared to the undoped or more than one v - doped clusters .
0.73 cm in order to understand the stability of the doped clusters , we have calculated the cohesive energies and the bond lengths of the co - co , co - v and v - v dimers as shown in table [ table6.1 ] .
it is seen that the v@xmath3 dimer is the most stable and the bond length of the v@xmath3 dimer is also about 14% shorter than that of the co@xmath3 dimer which has the least binding among the three . on the other hand , the cohesive energy and bond length of the co - v dimer
are intermediate of the co@xmath3 and the v@xmath3 dimers . for the co@xmath11v cluster
, we have seen that the center doping in icosahedral structure is the most favorable .
binary clusters are known to have the problem of segregation , where the doped atom can segregate at the surface or center . to a first approximation
, if one of the homonuclear bonds is the strongest , then that species tends to be at the center of the cluster.@xcite our finding of central v - doping is in accordance with this , as the v@xmath3 dimer has the strongest binding .
the atomic radius of a co - atom and a v - atom , being almost same , the substitution of the central co atom in an icosahedral co@xmath0 cluster by a v atom leaves the structure almost unperturbed . as for evidence , the center to vertex average distance and the average distance between two nearby vertices in the optimal icosahedral structure of the co@xmath0 cluster are 2.334 and 2.450 . for the optimal co@xmath11v cluster ,
these values are almost same , 2.344 and 2.465 respectively . therefore the large gain in cohesive energy of the co@xmath11v cluster over that of the co@xmath0 cluster
may be due to the enhanced cohesive energy of the cov dimer over that of the co@xmath3 dimer , though an atom in a dimer is quite different from an atom at the center of a 13-atom cluster .
the v atom being at the central position , there are maximum number of cov dimers in v - doped clusters .
0.2 cm .[table6.1]cohesive energies and bond lengths of co@xmath3 , co - v and v@xmath3 dimers in the present calculation . for comparison
, we have also listed the experimental values for co@xmath3 dimer ( ref .
35 ) and v@xmath3 dimer ( ref .
[ cols="^,^,^,^,^,^,^ " , ] we have then performed an exhaustive search for the mes , taking h@xmath3 at different possible places on the mes of the corresponding bare cluster for each composition . fig .
[ chemisorp ] shows our calculated lowest energy chemisorption structures after h@xmath3 absorbsion for clusters of all compositions .
first thing to notice is that h@xmath3 molecule chemisorbs dissociatively in each case , _
i.e _ the distance between the two h - atoms in the chemisorption structures is much larger than the h - h bond length of an isolated h@xmath3 molecule .
the chemisorption gives rise to moderate perturbation to the structures and makes them more symmetric ( _ i.e _ surface bonds are now less dispersive ) compared to the parent structure without hydrogen .
there are three possible ways for the h - atoms to be adsorbed on each cluster : on top of an atom ( one fold ) , at a bridge position between two atoms ( two fold ) and at the hollow site of a triangular plane on the cluster ( three fold ) .
again , as the surface contains two species of atoms for co@xmath26v@xmath3 , co@xmath27v@xmath28 and co@xmath32v@xmath33 , then the one fold position can be on top of a surface v - atom or on top of a surface co - atom .
two fold position can be the top of a v - v bond , co - v bond or a co - co bond which can be nearer to or away from the v - site .
similarly , in triangular plane , there are several possibilities : ( @xmath36 ) all the three atoms can be v - atoms ( only possible for co@xmath32v@xmath33 ) , ( @xmath37 ) one co atom and two v - atoms , ( @xmath38 ) two co - atoms and one v - atom , and ( @xmath39 ) all three are co - atoms . we have considered all these possible combinations during optimization .
it is , however , seen that in each of the optimized structures , h - atoms absorb at the hollow site on the surface and for m @xmath40 2 , they prefer the association of the local v atom .
for example in the co@xmath32v@xmath33h@xmath3 cluster , one h - atom absorbs at the hollow site of v - v - v triangular plane and the other h atom on top of a v - v - co triangular plane . for co@xmath26v@xmath3h@xmath3
, the h - atoms appear to absorb at the bridge positions , but they are inclined with an angle such that the absorption tends toward a three fold configuration .
the preference of more coordinated hollow site is likely due to geometric arrangements .
it allows the h - atom to interact more with the v or co atoms . on the other hand ,
the v - site preference of h - atom is due to the formation of stronger @xmath15-@xmath8 bond with v - atom compared to that with co - atom .
0.6 cm the higher reactivity of the co@xmath0 cluster compared to the co@xmath11v cluster , while both of them have the same twelve co - atoms on the surface , can be understood from charge transfer analysis .
we have computed the charge enclosed within a sphere around each surface atom of the optimal structures of co@xmath0h@xmath3 and co@xmath11vh@xmath3 clusters .
it is seen that the amount of charge on the surface co - atoms in the co@xmath0h@xmath3 cluster is larger by about 0.2 ( in unit of electronic charge ) than that of co@xmath11vh@xmath3 and it is mainly contributed by the @xmath8-electrons of surface co atoms
. therefore stronger 3@xmath8 - 1@xmath15 interaction between surface co - atoms and chemisorbed h - atoms in case of co@xmath0h@xmath3 increases its reactivity .
this result is in accordance with that of fujima _ et al _ @xcite who also predicted the stronger interaction between 1@xmath15 of h atom and the 3@xmath8 orbital of the surface co atom in the co@xmath0h@xmath3 cluster compared to that in the co@xmath11vh@xmath3 cluster using density of state analysis .
they showed that the anti - bonding orbital component between the h-1@xmath15 electron and the 3@xmath8 electron of the surface co - atoms shifts further away above the homo in case of co@xmath0h@xmath3 compared to that of co@xmath11vh@xmath3 , while the bonding orbital components for both the clusters stay almost at the same energy position below the homo . extending the charge transfer analysis for the co@xmath1v@xmath2h@xmath3 clusters with @xmath7 = 2,3 and 4
, we find that the average charge per surface v - atom is gradually increasing with the increase of v - concentration which indicates the formation of stronger v - h bonds at the surface with increasing v - doping .
average charge per surface co - atom for these three clusters , however , remains almost the same .
0.9 cm fig .
[ chemisorption ] shows the plot of chemisorption energy with increasing v - concentration , where the chemisorption energy is defined as @xmath41 it is a positive quantity and gives a measure of binding strength of h@xmath3 molecule with the cluster .
the plot shows a minimum for co@xmath11vh@xmath3 again indicating the lowest binding efficiency of co@xmath11v with h. however , chemisorption energy increases with increasing v concentration .
the source of this chemisorption energy is the cluster rearrangement energy ( _ i.e _ the energy change due to the geometrical rearrangement of the cluster upon chemisorption ) and the efficient cluster - absorbate bonding in presence of v due to more efficient @xmath15-@xmath8 hybridization .
to summarize , we have studied the geometric and electronic structures of v doped co@xmath0 clusters and their chemisorption towards hydrogen molecules using first principles density functional calculation .
the lowest energy structures of all compositions prefer to have icosahedral geometry , unlike hexagonal symmetry preference of the pure co clusters . for the co@xmath11v cluster ,
the single v atom prefers to sit at the central site , thereby guarded by all the surface co atoms and can not participate directly in the chemisorption reaction .
consequently reactivity of co@xmath11v becomes very less . on the other hand , for more than one v atom doping
, the additional v - atoms reside on the surface and come in direct contact with chemisorbed h - atom and reactivity increases .
our calculated spin gaps , density of states and charge transfer analysis explain nicely the stability of clusters and their tendency towards chemisorption . in the chemisorbed structures ,
h - atoms adsorb dissociatively at the more coordinated hollow sites and they prefer v - site association due to stronger 3@xmath8 - 1@xmath15 hybridization .
to have better insight into the chemisorption reaction , one needs to study the transition states for the optimal cluster of each composition and we believe that calculation of activation barriers will also lead to same conclusion as we have predicted here .
s. d. is thankful to council of scientific and industrial research ( government of india ) for financial support .
thanks department of science and technology , india for swarnajayanti fellowship and the support through advanced materials research unit .
r. ferrando , j. jellinek , and r. l. johnston , _ chem .
rev . _ * 108 * , 845 ( 2008 ) .
j. bansmann , s. h. baker , c. binns , j. a. blackman , j. -p .
bucher , j. dorantes - davila , v. dupuis , l. favre , d. kechrakos , a. kleibert , k. -h .
meiwes - broer , g. m. pastor , a. perez , o. toulemonde , k. n. trohidou , j. tuaillon , and y. xie , _ surf .
rep . _ * 56 * , 189 ( 2005 ) .
j. h. sinfelt , _ bimetallic catalysts : discoveries , concepts and allpications _ , j. wiley , new york , 1983 .
n. toshima , t. yonezawa , _ new j. chem . _
* 22 * , 1179 ( 1998 ) .
a. v. ruban , h. a. skriver , j. k. norskov , _ phys .
b _ * 59 * , 15990 ( 1999 ) .
g. bozzolo , j. ferrante , r. d. noebe , b. good , f. s. honecy , p. abel , _ comput .
sci . _ * 15 * , 169 ( 1999 ) .
f. besenbacher , i. chorkendorff , b. s. clausen , b. hammer , a. m. molenbroek , j. k. norskov , i. stensgaard , _ science _ * 279 * , 1913 ( 1998 ) .
a. m. molenbroek , s. haukka , b. s. clausen , _ j. phys
b _ * 102 * , 10680 ( 1998 ) .
a. bongiorno and u. landman , _ phys .
lett . _ * 95 * , 106102 ( 2005 ) .
e. kaxiras , _ phys
_ * 64 * , 551 ( 1990 ) .
g. h. guvelioglu , p. ma , x. he , r. c. forrey and h. cheng , _ phys . rev
. lett . _ * 94 * , 026103 ( 2005 ) ; y. jinlong , x. chuanyun , x. shangda and w. kelin , _ phys .
b _ * 48 * , 12155 ( 1993 ) ; r. q. zhang , t. s. chu , h. f. cheung , n. wang and s. t. lee , _ phys . rev . b _ * 64 * , 113304 ( 2001 ) ; j. gavnholt and j. schiotz , _ phys . rev .
b _ * 77 * , 035404 ( 2008 ) ; d. r. alfonso , s .- y .
wu , c. s. jayanthi , e. kaxiras , _ phys .
b _ * 59 * , 7745 ( 1999 ) ; s. k. nayak , s. e. weber , p. jena , k. wildberger , r. zeller , p. h. dederichs , v. s. stepanyuk and w. hergert , _ phys . rev .
b _ * 56 * , 8849 ( 1997 ) ; h. kawamura , v. kumar and y. kawazoe , _ phys . rev . b _ * 70 * , 193402 ( 2004 ) .
r. l. whetten , d. m. cox , d. j. trevor and a. kaldor , _ phys .
lett . _ * 54 * , 1494 ( 1985 ) ; l. holmgren , h. gronbeck , m. andersson and a. rosen , _ phys . rev .
b _ * 53 * , 16644 ( 1996 ) .
s. nonose , y. sone , k. onodera , s. sudo and k. kaya , _ j. phys . chem . _ * 94 * , 2744 ( 1990 ) .
g. c. bond , _ heterogeneous catalysis _
, 2nd ed . ;
clarendon press : oxford , 1987 . a. nakajima , t. kishi , t. sugioka , y. sone and k. kaya , _ j. phys . chem . _ * 95 * , 6833 ( 1991 ) .
a. pramann , k. koyasu and a. nakajima , _ j. phys
a _ * 106 * , 2483 ( 2002 ) .
r. l. whetten , d. m. cox , d. j. trevor and a. koldor , _ phys .
_ * 54 * , 1494 ( 1985 ) .
k. hoshino , t. naganuma , k. watanabe , y. konishi , a. nakajima , k. kaya , _ chem .
lett . _ * 239 * , 369 ( 1995 ) . vienna _
ab initio _ simulation package ( vasp ) , technische universitt wien , 1999 ; g. kresse and j. hafner , phys .
b * 47 * , 558 ( 1993 ) ; g. kresse and j. furthmuller , phys .
b * 54 * , 11169 ( 1996 ) .
p. e. blchl , phys .
b * 50 * , 17953 ( 1994 ) .
g. kresse and d. joubert , phys .
b * 59 * , 1758 ( 1999 ) .
j. p. perdew , k. burke and m. ernzerhof , phys .
lett . * 77 * , 3865 ( 1996 ) .
m. c. payne , m. p. teter , d. c. allan , t. a. arias and j. d. joannopoulos , _ rev .
phys . _ * 64 * , 1045 ( 1992 ) .
s. datta , m. kabir , s. ganguly , b. sanyal , t. saha - dasgupta and a. mookerjee , phys .
b , * 76 * , 014429 ( 2007 ) . ; ed .
j. jellinek , springer , berlin , p. 277
( 1999 ) and references therein .
j. jellinek , e. b. krissinel , _ chem .
lett . _ * 258 * , 283 ( 1996 ) ; j. jellinek , e. b. krissinel , _ chem .
_ * 272 * , 301 ( 1997 ) .
j. alonso , _ chem .
rev . _ * 100 * , 637 ( 2000 ) .
mukul kabir , d. g. kanhere and abhijit mookerjee , phys .
b * 73 * , 224443 ( 2006 ) ; t. m. briere , m. h. f. sluiter , v. kumar and y. kawazoe , phys . rev .
b * 66 * , 064412 ( 2002 ) .
o. dieguez , m. m. g. alemany , c. rey , p. ordejon and l. j. gallego , phys .
b * 63 * , 205407 ( 2001 ) . e. k. parks , l. zhu , j. ho and s. j. riley , j. chem . phys . *
102 * , 7377 ( 1995 ) ; n. n. lathiotakis , a. n. andriotis , m. menon and j. connolly , j. chem . phys . * 104 * , 992 ( 1996 ) .
s. datta , m. kabir , t. saha - dasgupta , and a. mookerjee ; to be published .
s. datta , ph .
d thesis , s. n. bose national centre for basic sciences ( 2008 ) .
x. xu , s. yin , r. moro and w. a. de heer , phys .
* 95 * , 237209 ( 2005 ) .
s. ganguly , m. kabir , s. datta , b. sanyal and a. mookerjee , phys .
b * 78 * , 014402 ( 2008 ) .
a. kant , b. strauss , _ j. chem . phys .
_ * 41 * , 3806 ( 1964 ) .
s. p. walch , c. w. bauschlicher jr . , b. o. roos and c. j. nelin , _ chem .
lett . _ * 103 * , 175 ( 1983 ) .
su , d. a. hales and p. b. amentrout , _ j. chem .
phys . _ * 99 * , 6613 ( 1993 ) .
m. el - batanouny , m. stronger , g. p. williams and j. colbert , _ phys .
_ * 46 * , 269 ( 1981 ) .
m. v. ruckman and m. strongin , _ phys .
b _ * 29 * , 7105 ( 1984 ) ; d. prigge , w. schlerk and e. bauer , _ surf .
lett . _ * 123 * , l698 ( 1982 ) .
g. kresse and j. hafner , _ surf .
_ * 459 * , 287 ( 2000 ) .
k. p. huber and g. hertzberg , _ molecular structure and molecular spectra iv : constants of diatomic molecules _
( van norstrand reinhold , new york , 1979 ) .
n. fujima and t. yamaguchi , _ j. phys .
society of japan _ * 61 * , 1724 ( 1992 ) . | structures , physical and chemical properties of v doped co@xmath0 clusters have been studied in detail using density functional theory based first - principles method .
we have found anomalous variation in stability of the doped clusters with increasing v concentration , which has been nicely demonstrated in terms of energetics and electronic properties of the clusters .
our study explains the nonmonotonic variation in reactivity of co@xmath1v@xmath2 clusters towards h@xmath3 molecules as reported experimentally [ j. phys .
chem . * 94 * , 2744 ( 1990 ) ] .
moreover , it provides useful insight into the cluster geometry and chemically active sites on the cluster surface , which can help to design better catalytic processes . |
information - theoretic properties are used in recent years for the study of quantum mechanical systems @xcite . in the present work
we carry out a systematic study of shannon information @xmath3 , onicescu information energy @xmath4 , order parameter @xmath5 and complexity @xmath6 , in atoms with @xmath7 . in previous work @xcite we proposed a universal property of @xmath3 for density distributions of nuclei , electrons in atoms and valence electrons in atomic clusters .
this property has the form @xmath8 where @xmath9 is the number of particles of the system and the parameters @xmath10 and @xmath11 depend on the system under consideration . recently @xcite we have obtained the same form for systems of correlated bosons in a trap . in the present paper
we employ very accurate spin - independent atomic wave functions obtained by bunge et al @xcite , by applying the roothaan - hartree - fock method ( rhf ) to calculate analytical self consistent - field atomic wave functions .
thus we verify the above relation for atoms , which was obtained in the past @xcite employing another set of electron wave functions .
thus we obtain a framework to be used as a basis for further work on information - theoretic properties of atoms .
an interesting question is the effect of the electron - electron correlations on @xmath3 and relation ( [ eq : ini ] ) , which could be answered in future research .
we focus our attention on the problem of similarity index based on the concept of an information distance .
the concept of similarity is an old one and related to the distinction between two or more objects @xcite .
specifically , in our paper we study two candidates connected with the concept of similarity or information distance .
the first one is the kullback - leibler relative entropy @xmath12 ( and also the symmetrized kullback distance @xmath13 ) and the second one is the jensen shannon divergence @xmath14 .
thus we are able to compare various density distributions obtained using various models .
it turns out that @xmath3 , @xmath13 and @xmath14 measures are useful for this purpose .
the framework developed in the present work for @xmath3 together with @xmath15 obtained previously @xcite with rigorous inequalities holding for @xmath3 , @xmath16 , @xmath17 enables us to calculate the so called complexity measure @xmath18 introduced in @xcite .
our procedure leads to the interesting result that complexity as function of @xmath0 shows shell effects at closed shells atoms i.e. for the above atoms complexity decreases with respect to neighboring atoms .
the outline of our paper is the following : in sec .
[ sec:2 ] we describe measures of information content of a quantum system together with information distances of two probability distributions and a complexity measure recently proposed . in sec . [ sec:3 ] we present our formalism , while sec .
[ sec:4 ] contains our numerical results and discussion .
[ sec:5 ] is a summary of the paper .
finally sec .
[ sec:6 ] contains our comments on the importance of our results on complexity .
the shannon information entropy @xcite @xmath16 in position - space may be defined as @xmath19 where @xmath20 is the electron density distribution normalized to unity .
the corresponding information entropy @xmath17 in the momentum space representation is @xmath21 where @xmath22 is the momentum density distribution normalized to unity studied in @xcite . in position - space @xmath16
determines the extent of electron delocalization , since it tends to a maximum as the distribution flattens out and deviates from this maximum when structure is introduced in @xmath20 . in momentum - space
a maximum in @xmath17 corresponds to a delocalized distribution in momentum - space @xcite .
the total information entropy is given by @xmath23 where @xmath3 , @xmath16 and @xmath17 obey the following rigorous inequalities @xcite @xmath24 the lower and the upper limits can be written , for density distributions normalized to one @xmath25 where @xmath26 is the mean square radius and @xmath27 is the kinetic energy .
another measure of information content of a quantum system is the concept of information energy @xmath4 introduced by onicescu in an attempt to define a finer measure of dispersion distribution than that of shannon information entropy @xcite . for a discrete probability distribution @xmath28
, @xmath4 is defined as @xmath29 which is extended for a continuous density distribution @xmath30 as @xmath31 so far , only the mathematical aspects of the concept have been developed , while the physical aspects have been ignored .
a recent study of @xmath4 for atomic nuclei has been carried out in @xcite .
the meaning of ( [ eq : onicont ] ) can be seen by the following simple argument : for a gaussian distribution of mean value @xmath32 , standard deviation @xmath33 and normalized density @xmath34\ ] ] relation ( [ eq : onicont ] ) gives @xmath35\,dx=\frac{1}{2\sigma\sqrt{\pi}}\ ] ] @xmath4 is maximum if one of the @xmath36 equals 1 and all the others are equal to zero i.e. @xmath37 , while @xmath4 is minimum when @xmath38 , hence @xmath39 ( total disorder ) .
the fact that @xmath4 becomes minimum for equal probabilities ( total disorder ) , by analogy with thermodynamics , it has been called information energy , although it does not have the dimension of energy @xcite .
it is seen from ( [ eq : oninorm ] ) that the greater the information energy , the more concentrated is the probability distribution , while the information content decreases .
thus one can define a measure of information content analogous to shannon s @xmath3 by the relation @xmath40 relation ( [ eq : onicont ] ) is extended for a 3-dimensional spherically symmetric density distribution @xmath41 : @xmath42 in position and momentum space respectively , where @xmath43 is the corresponding density distribution in momentum space .
@xmath44 has dimension of inverse volume , while @xmath45 of volume .
thus the product @xmath46 is dimensionless and can serve as a measure of concentration ( or information content ) of a quantum system .
it is also seen from ( [ eq : oninorm ] ) that @xmath4 increases as @xmath33 decreases ( or concentration increases ) and shannon s information entropy ( or uncertainty ) @xmath3 decreases .
thus @xmath3 and @xmath4 are reciprocal . in order to be able to compare them
, we redefine the quantity @xmath47 by @xmath48 as a measure of the information content of a quantum system in both position and momentum spaces .
landsberg @xcite defined the order parameter @xmath5 ( or disorder @xmath49 ) as @xmath50 where @xmath3 is the information entropy ( actual ) of the system and @xmath15 the maximum entropy accessible to the system .
thus the concepts of entropy and disorder are decoupled and it is possible for the entropy and order to increase simultaneously . it is noted that @xmath51 corresponds to perfect order and predictability , while @xmath52 means complete disorder and randomness . in @xcite a measure of complexity @xmath18 was defined of the form @xmath53 which is called the `` simple complexity of disorder strength @xmath54 and order strength @xmath55 '' . when @xmath56 and @xmath57 `` complexity '' is an increasing function of `` disorder '' , and we have a measure of category i ( fig.1 of @xcite ) .
when @xmath58 and @xmath59 , `` complexity '' is an increasing function of `` order '' and we have a measure of category iii .
when both @xmath54 and @xmath55 are nonvanishing and positive ( @xmath57 , @xmath59 ) , `` complexity '' vanishes at zero `` disorder '' and zero `` order '' and has a maximum of @xmath60 at @xmath61 and @xmath62 .
this is complexity of category ii according to @xcite .
several cases for both @xmath54 and @xmath55 non - negative are shown in fig.2 of @xcite where @xmath18 is plotted as function of @xmath49 . in the present work we can find @xmath63 or @xmath64 as function of @xmath0 .
thus we are able to plot the dependence of @xmath18 on the atomic number @xmath0 .
the kullback - leibler relative information entropy @xmath12 @xcite for any probability distributions @xmath65 is defined by @xmath66 which for continuous probability distributions is defined as @xmath67 ( [ eq : eq6 ] ) , ( [ eq : eq7 ] ) can be easily extended for 3-dimensional systems .
@xmath12 measures the difference of distance of @xmath68 from the reference ( or a priori ) distribution @xmath69 .
it satisfies : @xmath70 for any distributions @xmath71 .
it is a measure which quantifies the distinguishability ( or distance ) of @xmath68 from @xmath69 , employing a well - known concept in standard information theory . in other words
it describes how close is @xmath68 to @xmath69 by carrying out observations or coin tossing , namely trials @xmath72 ( in the sense described in @xcite ) .
however , the distance @xmath12 does not satisfy the triangle inequality and in addition is i ) not symmetric
ii ) unbounded and iii ) not always well defined @xcite . to avoid these difficulties rao and lin @xcite introduced a symmetrized version of @xmath12 @xcite , the jensen - shannon divergence @xmath73 where @xmath74 stands for shannon s entropy
@xmath14 is minimum for @xmath75 and maximum when @xmath76 and @xmath77 are two distinct distributions , when @xmath78 . in our case
@xmath14 can be easily generalized for continuous density distributions .
for @xmath14 minimum the two states represented by @xmath76 and @xmath77 are completely indistinguishable , while for @xmath14 maximum they are completely distinguishable .
the amount of distinguishability can be further examined by using wooters criterion @xcite .
two probability distributions @xmath76 and @xmath77 are distinguishable after @xmath72 trials @xmath79 if and only if @xmath80 .
the present work is a first step to examine the problem of comparison of probability distributions ( for atoms and various models ) which is an area well developed in statistics , known as information geometry @xcite .
the key quantities in our work are the density distribution @xmath81 and the momentum distribution @xmath82 . in general the calculation of @xmath81 and @xmath82
presuppose the knowledge of the one body density matrix @xmath83 which is defined as @xmath84 where @xmath85 is the wave function that describes the system under consideration ( in the present work the atom ) .
@xmath20 is just the diagonal part of the @xmath86 , i.e. @xmath87 while @xmath22 is the fourier transform of the one - body density matrix i.e. @xmath88 d{\bf r}_1 ' d{\bf r}_1\ ] ] in the framework of the hartree - fock approximation , which is applied in the present work , @xmath85 has the well known form of a slater determinant ( electrons of the atoms consist a system which obeys the fermi - dirac statistics ) .
it is easy to prove that in that case @xmath86 takes the following simple form @xmath89 where @xmath90 is the single particle wave function describing the electrons in an atom .
the index @xmath91 runs over all @xmath0 single particle states .
now @xmath81 and @xmath82 are written as @xmath92 @xmath93 where @xmath94 the wave functions @xmath90 and @xmath95 are decomposed in the usual form @xmath96 and @xmath97 the radial momentum wave function @xmath98 is related to the radial wave function in coordinate space through @xmath99 where @xmath100 is a spherical bessel function . in the present work we consider very accurate spin - independent atomic wave functions obtained by bunge _
et al _ @xcite by applying the roothaan - hartree - fock method to calculate analytical self - consistent - field atomic wave function . in this approach
the radial atomic orbitals @xmath101 are expanded as a finite superposition of primitive radial functions @xmath102 where the normalized primitive basis @xmath103 is taken as a slater - type orbital set , @xmath104 where the normalization factor @xmath105 is given by @xmath106^{1/2 } \label{}\ ] ] and @xmath107 is the principal quantum number , @xmath108 is the orbital exponent , and @xmath109 is the azimuthal quantum number . in general @xmath81 is not spherically symmetric and depends in addition on the angle @xmath110 . in the present work we consider @xmath81 averaged spherically ( the same holds also for @xmath82 ) .
the shannon information entropy in position - space , momentum space and in total are given in eqs .
( [ eq : sr ] ) , ( [ eq : sk ] ) and ( [ eq : stot ] ) respectively .
another quantity which gives information about the localization ( delocalization ) of the atomic systems is the local shannon entropy defined ( in position- and momentum - space respectively ) as follows @xmath111 in order to formulate the concept of the similarity or information distance between two atomic systems with @xmath112 and @xmath113 the corresponding density distributions , the kullback - leibler relative entropy is defined as @xmath114 which may be interpreted as a measure of deviation of @xmath115 from @xmath113 .
the corresponding symmetrized kullback distance @xmath13 is @xmath116 the physical meaning of the kullback distance is very clear . @xmath12 is equal to zero for two identical species and approaches infinity as the difference between @xmath112 and @xmath113 increases . @xmath12 and @xmath13 in momentum space are defined using @xmath117 and @xmath118 in the same way .
the jensen - shannon divergence entropy is defined as follows @xmath119 the physical meaning of @xmath14 is similar to @xmath13 , while the definition in momentum space is defined using @xmath82 in the same way .
another aspect of our work is to compare rhf densities with those of the work @xcite in the framework of the @xmath13 and @xmath14 .
more specifically we compare the density distributions of the electrons originating from the paper of bunge et.al .
@xcite ( used in the present work ) with the phenomenological one of sagar et.al .
@xcite who employed the phenomenological form @xmath120 where @xmath121 is the first ionization potential of the system .
we also compare rhf densities with those of the well known thomas - fermi model . for the density distribution we use the simple form obtained by sommerfeld @xcite where @xmath122 where @xmath123 , @xmath124 , @xmath125 and @xmath126 .
the normalization constant @xmath127 is calculated from the normalization condition @xmath128
in fig . [ fig:1 ] ( a ) and ( b ) we plot the shannon information entropy both in coordinate - space ( @xmath16 ) and momentum - space ( @xmath17 ) as functions of the electron number @xmath0 . in @xmath16
coexist an average increasing behavior and also an obvious shell effect structure around fully filled shells , such as he , ne , ar , kr where there are minima of the curve @xmath129 .
the physical meaning of that behavior is that @xmath41 for these atoms is the most compact one when compared to the neighboring atoms .
the values of @xmath17 ( fig .
[ fig:1 ] ( b ) ) show a monotonic increase with @xmath0 .
however there is also in the behavior of @xmath17 a local shell effect . in fig .
[ fig:2 ] ( a ) we plot the total shannon information entropy @xmath3 .
@xmath3 is a strictly monotonic increasing function of @xmath0 with only two exceptions for ni and pd .
these exceptions are due to the fact that @xmath16 and @xmath17 depend on the arrangement of the electrons in shells .
there is a delicate balance between @xmath16 and @xmath17 resulting in the general rule that @xmath130 is a monotonic increasing function of @xmath0 except in ni and pd where the electron arrangement in shells is such that the decrease of the value of @xmath16 can not be balanced by a corresponding increase of @xmath17 .
thus the monotonicity of @xmath3 is slightly violated .
a shell effect is also obvious in the behavior of @xmath3 i.e. minima at closed shells .
[ fig:2 ] ( b ) illustrates the trend of @xmath3 as a function of @xmath131 .
the best linear fit is also plotted in the same figure where @xmath132 .
it is noted that this result is not new but has been already obtained using other wave functions in @xcite . in the present work
we verify this result with rhf electron wave functions @xcite and we employ this framework for new calculations . in fig .
[ fig:3 ] ( a ) , ( b ) , ( c ) , ( d ) we plot the complexity measure @xmath18 in atoms for various values of parameters @xmath54 and @xmath55 .
it is seen that for all sets of @xmath54 and @xmath55 , @xmath18 shows qualitatively the same trend as function of @xmath0 i.e. it fluctuates around an average value and shows local minima for atoms with closed shells .
these results compare favorably with intuition i.e. complexity is less at closed shells which can be considered more compact than neighboring nuclei and consequently less complex .
it is noted that this result comes from a procedure which is not trivial i.e. first we calculate @xmath81 and @xmath82 , second we find @xmath130 from the shannon definition and @xmath15 employing rigorous inequalities and third we obtain the complexity measure introduced in @xcite .
[ fig:4 ] ( a ) displays the values of the onicescu information content @xmath47 versus @xmath0 .
the first three maxima correspond to the fully closed shells ( he , ne and ar ) where in the case of the next closed shell ( kr ) a local maximum exists . in fig .
[ fig:4 ] ( b ) we plot in the same footing the onicescu information content @xmath47 and the ionization potential @xmath121 .
it is indicated that @xmath47 and @xmath121 are correlated in the sense that there is a similarity in the trend of values of @xmath47 and @xmath121 as functions of @xmath0 .
this similarity is more obvious in regions of small @xmath0 where linear relations @xmath133 can be extracted for regions li - ne and na - ar .
however , it seems that , there is no universal relation between them .
there are many entropic measures of spread of probability densities e.g. @xmath3 , @xmath4 e.t.c .
but researchers prefer @xmath3 because of its unique properties , while @xmath4 was introduced by onicescu as a sensitive measure of information .
however , @xmath3 and @xmath4 are different functionals of the density and their relation is difficult to find . in fig .
[ fig:5 ] ( a ) we display the symmetrized kullback distance between the rhf density distributions and the approximate one given by the relation ( [ d - sag-1 ] ) .
it is obvious that the symmetrized kullback distance @xmath13 becomes minimum in the case of the fully closed shells atoms .
this physically means that the approximate @xmath81 ( [ d - sag-1 ] ) works better in closed shell atoms than the open shells .
in addition there is an increasing trend with @xmath0 .
that means that in general @xmath134 ( [ d - sag-1 ] ) fails to describe the structure of heavier atoms .
[ fig:5 ] ( b ) displays the jensen - shannon divergence entropy @xmath14 as a function of @xmath0 .
the behavior is almost similar to @xmath13 and the comments are the same . in fig .
[ fig:5 ] ( c ) we display @xmath14 versus @xmath13 .
there is a linear relation of the two information distances . in fig .
[ fig:6 ] ( a ) we display @xmath13 as an information distance between the rhf density distribution and the thomas - fermi density distribution given by the relation ( [ tf - dd ] ) .
it is seen that the values of @xmath13 decrease with @xmath0 .
this is expected because the thomas - fermi approximation works well in heavier atoms .
so the two distributions are closer in heavier atoms than in the light ones . the local maxima of the values of @xmath13 correspond to the closed shell atoms .
this deviation is due to the absence of shell effect character of the thomas - fermi distribution compared to the realistic rhf density distribution . fig .
[ fig:6 ] ( b ) displays @xmath14 as a function of @xmath0 .
the behavior is almost similar to @xmath13 . in fig .
[ fig:6 ] ( c ) we display @xmath14 versus @xmath13 . the local shannon entropy ( [ eq : localshannon ] ) both in position- and momentum - space is presented in fig . [ fig:7 ] for various atoms .
as it is pointed out in @xcite in systems with density @xmath135 near the nucleus the local shannon entropy in position - space @xmath136 will be negative , thus the contribution to the integral from this region will serve to lower its value ( localization ) while in region where @xmath137 , such as the valence , contribution to the shannon entropy in position - space will be positive which leads to delocalization .
in contrast the local shannon entropy in momentum - space @xmath138 is always positive .
this is due to the fact that @xmath139 for all @xmath140 .
so , there is no negative contribution to the shannon entropy in momentum - space .
it is worth , discussing the behavior of the momentum distribution and as a consequence the local shannon entropy , to mention that the precise knowledge of the electron momentum distribution is important for atoms used as a dark matter or neutrino detectors .
in such a kind of experiments the single particle wave function in momentum space or in general the momentum density of the electrons are the main ingredient of the relative cross sections .
the trend ( localization , delocalization , etc ) of the electron wave function or momentum distribution affect considerably the values of the cross sections , especially in experiments where the production of electrons in neutralino - nucleus or neutrino - nucleus collisions are investigated @xcite .
finally in table [ tab:1 ] we tabulate for the sake of reference the quantities @xmath16 , @xmath17 , @xmath3 , @xmath15 , @xmath5 , and @xmath47 for each atom as functions of @xmath0 .
we include the results for hydrogen which are known exactly i.e. @xmath141 and @xmath142 .
in previous works the universal relation @xmath143 was proposed for the information entropy @xmath3 as function of the number of particles @xmath9 in atoms , nuclei , atomic clusters and correlated bosons in a trap i.e. systems of various sizes , with various interactions , obeying different statistics ( fermions and bosons ) . in this paper
we verify the above relation employing rhf electron density distributions for atoms .
thus we construct a basis in order to study some information - theoretic properties of atoms .
specifically , we calculate the symmetrized kullback - leibler relative entropy @xmath13 and the jensen - shannon divergence @xmath14 which serve as measures of information distance of probabilities distributions and are useful to compare electron distributions according to various models .
two examples are given .
we compare rhf density distributions first with an asymptotic density depending on the ionization potential and second with the well - known thomas fermi approximation .
we also obtain onicescu s information energy @xmath4 and its corresponding information measure , which correlates with the ionization potential .
finally , we calculate a recently proposed complexity measure @xmath18 inspired by landsberg s order parameter @xmath5 .
it turns out that the function @xmath144 shows the interesting feature that for closed shells atoms is smaller than neighboring ones .
this indicates that closed shells atoms are less complex than neighboring ones , which compares favorably with expectations according to intuition .
there is a long debate in the literature on order , disorder , complexity and organization for physical , biological and other systems .
a generally accepted quantitative definition of complexity does not exist so far .
more work has been carried out in classical systems and much less in quantum ones .
our aim is to contribute in this debate examining complexity in atoms ( for the first time to our knowledge ) adopting a particular definition of complexity measure according to shiner , davison and landsberg @xcite .
the reason for our choice is that we can easily calculate it for atoms ( and we intend to calculate it for other quantum systems as well ) , because our previous experience with information entropy facilitates this .
more work is needed in the future by using the so called statistical measure of complexity of lopez - ruiz , manchini and calbet @xcite in
@xcite the authors studied disorder and complexity in an ideal fermi gas of electrons .
they observed that for a small number of electrons landsberg s order parameter @xmath5 is small , while @xmath5 increases as one pumps electrons into the system and the energy levels fill up .
this result is in a way counter - intuitive and indicates that as particles are added in a quantum - mechanical system , the system becomes more ordered .
this result was further corroborated in @xcite by calculating @xmath5 as function of the number of particles @xmath9 for realistic quantum systems i.e. atomic nuclei and atomic clusters . in @xcite @xmath145
was obtained for bosonic systems as well ( correlated atoms in a trap ) .
all cases show the same trend for @xmath145 i.e. @xmath5 is an increasing function of @xmath9 . however , if one is interested how complex ( or organized ) is a system , the information entropy @xmath3 ( sometimes used as a measure of disorder ) or the order parameter @xmath5 ( or disorder @xmath146 ) are not suitable measures of complexity or organization .
we note that these terms are interrelated from a semantic point of view
. @xmath18 has the advantage of being a convex measure of order ( or disorder ) i.e. it vanishes for highly ordered and disordered systems . in other words , it satisfies the `` one - hump '' criterion for statistical complexity measures .
thus @xmath18 is obtained by multiplying a measure of `` order '' by a measure of `` disorder '' .
a simple example is a perfect crystal which has perfect order and an ideal gas with complete disorder . both have zero complexity and fit well with the definition of @xmath18 . in the present paper
we find that complexity is less at closed shell atoms .
this satisfies our intuition , at least does not contradict common sense and indicates that our procedure from electron densities to information entropy and maximum entropy to complexity measure @xmath18 is reasonable . the fact that @xmath18 fluctuates around an average value is new and interesting , because it shows that complexity of atoms does not increase as the atomic number @xmath0 increases . in other words , as one pumps electrons into the atom , the atom has not the ability to grow in complexity .
the question is open what happens if atoms form molecules , molecules form more complex systems e.t.c .
we mention that the question whether physical or biological systems have the ability for organized complexity without the intervention of an external factor or agent is a hot subject in the community of scientists interested in complexity and can be extrapolated even in philosophical questions .
the work of k. ch .
chatzisavvas and c. p.panos was supported by herakleitos research scholarships ( 21866 ) of @xmath147 and the european union while the work of ch . c. moustakidis was supported by the pythagoras ii research project ( 80861 ) of @xmath147 and the european union .
the authors would like to thank prof . c. f. bunge for kindly providing the data for the atomic wave functions , prof .
s. e. massen for useful comments on the manuscript and also prof r. p. sagar for valuable comments and correspondence .
+ @xmath0 & atom & @xmath16 & @xmath17 & & & @xmath5 & + 1 & h & 4.14473 & 2.42186 & 6&56659 & 7&9054 & 0.21113 & 120&26700 + 2 & he & 2.69851 & 3.91342 & 6&61193 & 7&0493 & 0.06204 & 100&36100 + 3 & li & 3.07144 & 3.99682 & 7&69826 & 10&3578 & 0.25677 & 9&15713 + 4 & be & 3.62386 & 4.19019 & 7&81405 & 10&3950 & 0.24829 & 8&45434 + 5 & b & 3.40545 & 4.70590 & 8&11135 & 10&3738 & 0.21810 & 15&96530 + 6 & c & 3.10602 & 5.15658 & 8&26260 & 10&2624 & 0.19492 & 25&71210 + 7 & n & 2.80169 & 5.54934 & 8&35103 & 10&1520 & 0.17740 & 37&43200 + 8 & o & 2.55054 & 5.86737 & 8&41791 & 10&1113 & 0.16747 & 48&48340 + 9 & f & 2.29883 & 6.16333 & 8&46215 & 10&0533 & 0.15827 & 61&14500 + 10 & ne & 2.05514 & 6.43707 & 8&49221 & 9&9908 & 0.14999 & 75&24470 + 11 & na & 2.33009 & 6.48310 & 8&81319 & 11&6463 & 0.24326 & 15&86900 + 12 & mg & 2.39540 & 6.51440 & 8&91038 & 11&8296 & 0.24677 & 10&19480 + 13 & al & 2.44569 & 6.61928 & 9&06497 & 12&0615 & 0.24843 & 12&76590 + 14 & si & 2.41914 & 6.73380 & 9&15294 & 12&0500 & 0.24042 & 15&63600 + 15 & p & 2.35903 & 6.84865 & 9&20767 & 11&9954 & 0.23240 & 18&62490 + 16 & s & 2.29932 & 6.94939 & 9&24871 & 11&9769 & 0.22779 & 20&80380 + 17 & cl & 2.22174 & 7.05243 & 9&27418 & 11&9315 & 0.22716 & 23&36830 + 18 & ar & 2.13383 & 7.15541 & 9&28924 & 11&8758 & 0.21780 & 26&21000 + 19 & k & 2.30177 & 7.17242 & 9&47419 & 12&9220 & 0.26682 & 10&37540 + 20 & ca & 2.36309 & 7.18250 & 9&54334 & 13&0994 & 0.27147 & 6&66998 + 21 & sc & 2.29814 & 7.30329 & 9&60143 & 13&0334 & 0.26332 & 8&12090 + 22 & ti & 2.21855 & 7.42693 & 9&64548 & 12&9721 & 0.25644 & 9&48793 + 23 & v & 2.13512 & 7.54717 & 9&68229 & 12&9160 & 0.25036 & 10&91540 + 24 & cr & 1.95589 & 7.75135 & 9&70724 & 12&5913 & 0.22905 & 27&02960 + 25 & mn & 1.96257 & 7.77688 & 9&73945 & 12&8162 & 0.24007 & 13&94220 + 26 & fe & 1.88213 & 7.88265 & 9&76478 & 12&7697 & 0.23532 & 15&78050 + 27 & co & 1.80001 & 7.98612 & 9&78613 & 12&7264 & 0.23104 & 17&65210 + 28 & ni & 1.71826 & 7.99294 & 9&71120 & 12&7198 & 0.23653 & 18&83670 + 29 & cu & 1.56325 & 8.25076 & 9&81401 & 12&4646 & 0.21265 & 41&47510 + 30 & zn & 1.55625 & 8.27867 & 9&83493 & 12&6106 & 0.22011 & 23&74300 + 31 & ga & 1.57444 & 8.32388 & 9&89832 & 12&8676 & 0.23076 & 23&52780 + 32 & ge & 1.56746 & 8.37150 & 9&93896 & 12&9098 & 0.23012 & 24&53250 + 33 & as & 1.54980 & 8.41828 & 9&96808 & 12&9082 & 0.22777 & 25&68280 + 34 & se & 1.53425 & 8.45851 & 9&99276 & 12&9298 & 0.22715 & 25&81220 + 35 & br & 1.51064 & 8.49958 & 10&01020 & 12&9230 & 0.22540 & 26&44050 + 36 & kr & 1.48146 & 8.54092 & 10&02240 & 12&9026 & 0.22323 & 27&33700 + 37 & rb & 1.57626 & 8.54430 & 10&12060 & 13&7376 & 0.26329 & 12&45970 + 38 & sr & 1.62144 & 8.54228 & 10&16370 & 13&9159 & 0.26963 & 7&91895 + 39 & y & 1.61438 & 8.59046 & 10&20480 & 13&8772 & 0.26463 & 9&57134 + 40 & zr & 1.59462 & 8.64054 & 10&23520 & 13&8304 & 0.25995 & 10&94010 + 41 & nb & 1.52486 & 8.73236 & 10&25720 & 13&5780 & 0.24457 & 23&59820 + 42 & mo & 1.49117 & 8.78233 & 10&27350 & 13&5247 & 0.24039 & 25&89550 + 43 & tc & 1.50762 & 8.79074 & 10&29840 & 13&7087 & 0.24877 & 14&73730 + 44 & ru & 1.40305 & 8.87561 & 10&30610 & 13&4876 & 0.23588 & 28&97120 + 45 & rh & 1.39635 & 8.92206 & 10&31840 & 13&4670 & 0.23380 & 30&46670 + 46 & pd & 1.30482 & 8.98908 & 10&29390 & 13&1570 & 0.21761 & 47&72420 + 47 & ag & 1.32346 & 9.01391 & 10&33740 & 13&4264 & 0.15816 & 33&39990 + 48 & cd & 1.33132 & 9.02613 & 10&35740 & 13&5464 & 0.23541 & 21&51740 + 49 & in & 1.35191 & 9.04892 & 10&40080 & 13&7432 & 0.24320 & 20&89160 + 50 & sn & 1.35701 & 9.07295 & 10&43000 & 13&7921 & 0.24377 & 21&00720 + 51 & sb & 1.35483 & 9.09683 & 10&45170 & 13&8039 & 0.24285 & 21&24880 + 52 & te & 1.35345 & 9.11719 & 10&47060 & 13&8313 & 0.24297 & 20&80400 + 53 & i & 1.34658 & 9.13832 & 10&48490 & 13&8349 & 0.24214 & 20&74270 + 54 & xe & 1.33582 & 9.16022 & 10&49600 & 13&8264 & 0.24087 & 20&89380 + and ( b ) in momentum - space @xmath17 , as a function of the electron number @xmath0.[fig:1],title="fig:",width=226,height=264 ] + and ( b ) in momentum - space @xmath17 , as a function of the electron number @xmath0.[fig:1],title="fig:",width=226,height=264 ] versus @xmath0 , ( b ) onicescu information content @xmath47 and ionization potential @xmath121 ( in hartree units ) versus @xmath0.[fig:4],title="fig:",width=226,height=264 ] + versus @xmath0 , ( b ) onicescu information content @xmath47 and ionization potential @xmath121 ( in hartree units ) versus @xmath0.[fig:4],title="fig:",width=226,height=264 ] ) ) ( b ) corresponding jensen - shannon divergence entropy versus @xmath0 ( c ) jensen - shannon divergence versus the symmetrized kullback distance.[fig:5],title="fig:",width=226,height=264 ]
+ ) ) ( b ) corresponding jensen - shannon divergence entropy versus @xmath0 ( c ) jensen - shannon divergence versus the symmetrized kullback distance.[fig:5],title="fig:",width=226,height=264 ] + ) ) ( b ) corresponding jensen - shannon divergence entropy versus @xmath0 ( c ) jensen - shannon divergence versus the symmetrized kullback distance.[fig:5],title="fig:",width=226,height=264 ] ( b ) corresponding jensen - shannon divergence entropy versus @xmath0 ( c ) jensen - shannon divergence versus the symmetrized kullback distance.[fig:6],title="fig:",width=226,height=264 ] + ( b ) corresponding jensen - shannon divergence entropy versus @xmath0 ( c ) jensen - shannon divergence versus the symmetrized kullback distance.[fig:6],title="fig:",width=226,height=264 ] + ( b ) corresponding jensen - shannon divergence entropy versus @xmath0 ( c ) jensen - shannon divergence versus the symmetrized kullback distance.[fig:6],title="fig:",width=226,height=264 ] and ( b ) in momentum - space @xmath138 , as a function of the electron number @xmath0.[fig:7],title="fig:",width=226,height=264 ] + and ( b ) in momentum - space @xmath138 , as a function of the electron number @xmath0.[fig:7],title="fig:",width=226,height=264 ] | shannon information entropies in position and momentum spaces and their sum are calculated as functions of @xmath0 ( @xmath1 ) in atoms .
roothaan - hartree - fock electron wave functions are used .
the universal property @xmath2 is verified .
in addition , we calculate the kullback - leibler relative entropy , the jensen - shannon divergence , onicescu s information energy and a complexity measure recently proposed .
shell effects at closed shells atoms are observed .
the complexity measure shows local minima at the closed shells atoms indicating that for the above atoms complexity decreases with respect to neighboring atoms .
it is seen that complexity fluctuates around an average value , indicating that the atom can not grow in complexity as @xmath0 increases .
onicescu s information energy is correlated with the ionization potential .
kullback distance and jensen - shannon distance are employed to compare roothaan - hartree - fock density distributions with other densities of previous works . |
models for growing cells generally consist of substrates(@xmath0 ) and active proteins that catalyze their own synthesis and that of other components .
for example , in the models developed by scott _
et al._@xcite and maitra _
et al._@xcite , the active proteins correspond to ribosomes .
this class of models involving catalytic proteins can be used to accurately describe the exponential growth of a cell under a sufficient supply of substrates ; however , once the degradation rate of the active protein exceeds its rate of synthesis under a limited substrate supply , the cell s volume will shrink , leading to cell death .
hence , a cell population either grows exponentially or dies out , and in this cellular state it is not possible to maintain the population without growth .
+ to model a state with such suppressed growth , we consider two more chemical species , inhibitors(@xmath1 ) and active protein - inhibitor complexes(@xmath2 ) , in addition to the substrates(@xmath0 ) and active proteins(@xmath3 ) that are commonly adopted in models of cell growth . a schematic representation of the present model is shown in fig.[fig : fig1].(a ) . here
, we focus on two classes of proteins that are essential to the description of cellular growth : an active protein and inhibitor .
the active proteins are those that catalyze their own growth such as ribosomes , and can include metabolic enzymes , transporters , and growth - facilitating factors .
inhibitory proteins form a complex with active proteins , thereby suppressing their catalytic synthesis function
. they can be inhibitory factors such as yfia or hpf in _
other candidates for such inhibitors are misfolded or mistranslated proteins that are produced erroneously during the replication of active proteins , which inhibit the catalytic activity of active proteins by trapping them into the aggregates of misfolded proteins @xcite . our model , then ,
is given by @xmath4 where @xmath5 and @xmath6 represent the synthesis rate of the active protein and inhibitor , to be presented below .
@xmath7 denotes the reaction of complex formation , given by @xmath8 , @xmath9 represents the external concentration of substrate , @xmath10 denotes the spontaneous degradation rate of macromolecules @xmath11 , and @xmath12 represents the specific growth rate of the cell , given by @xmath13 .
+ in this model , the cell takes up substrates from the external environment from which active proteins and inhibitors are synthesized .
these syntheses @xmath14 , and @xmath15 , as well as the uptake of substrates , take place with the aid of catalysis by the active proteins . then , by assuming that the synthesized components are used for growth in a sufficiently rapid period , the growth rate is set to be proportional to the rate of active protein synthesis .
next , the catalytic activity of the active protein is inactivated due to the formation of an active protein - inhibitor complex @xmath16 , which , for example , corresponds to the interaction between a ribosome and yfia and hpf @xcite .
all chemical components are diluted by the volume growth of a cell , although they are spontaneously degraded at a much lower rate .
the complex has higher stability than active protein and inhibitor , alone ( @xmath17 is smaller than @xmath18 and @xmath19 ) .
+ it has been well established that inhibitory factors are actively synthesized under a resource - limited condition @xcite . moreover , interpreting @xmath1 molecules as incorrect polymers , they are expected to increase with the decrease of supply substrates , since this situation will limit the proofreading mechanism to eliminate them@xcite .
thus , with this interpretation of the inhibitors as incorrect polymers and also in consistency with increase in inhibitory factors under resource - limited condition , it naturally follows that the ratio of the synthesis of active protein to inhibitors is an increasing function of substrate concentration , i.e. , @xmath20 . in the model
, we assume that this ratio increases with the concentration and becomes saturated at higher concentrations , as in michaelis - menenten s form , and choose @xmath21 and @xmath22 , for example .
( see also the _ supplementary information _ for the derivation of such form in the case of a proofreading mechanism ) .
+ note that by summing up @xmath23 and @xmath24 , we obtain @xmath25 if @xmath18 and @xmath17 are zero ( or negligible ) .
it means that if the cell once reaches any steady state , the relationship @xmath26 is kept satisfied as long as @xmath3 and @xmath5 are not zero .
we use the relationship and eliminate @xmath2 by substituting @xmath27 for analysis below .
the steady state of the present model exhibits three distinct phases as a function of the external substrate concentration @xmath28 ( fig.[fig : fig1].(b ) ) , as computed by its steady - state solution .
the three phases are distinguished by both the steady growth rate and the concentration of active protein , which are termed as the active , inactive , and death phases , as shown in the figure , whereas the growth rate shows a steep jump at the boundaries of the phases .
the phases are characterized as follows . *
( i ) * in the active phase , the highest growth rate is achieved , where abundant active proteins work freely as catalysts . *
( ii ) * in the inactive phase , growth rate is not zero but is drastically reduced with orders of magnitude compared with the active phase . here ,
almost all active proteins are arrested by complex formation with the inhibitor , and their catalytic activity is deactivated . *
( iii ) * at the death phase , a cell can not grow , and all of the active proteins , inhibitors , and complexes go to zero . in this case
, the cell goes beyond the so - called `` point of no return '' and can never grow again , regardless of the amount of increase in @xmath28 , since the catalysts are absent in any form .
( as will be shown below , the active and inactive phases correspond to the classic log and stationary phases , but to emphasize the single - cell growth mode , we adopt these former terms for now ) . + the transition from the active to inactive phase is caused by the interaction between the active protein and inhibitor . in the substrate - poor condition
, the amount of inhibitor greatly exceeds the total amount of catalytic proteins ( @xmath29 ) , and any free active protein remaining vanishes . below the transition point from the inactive to death phase , the spontaneous degradation rate surpasses the synthesis rate , at which point all of the components decrease . this transition point
is simply determined by the balance condition @xmath30 .
hence , if @xmath18 is set to zero , the inactive - death transition does not occur .
+ we now consider the time series of biomass ( the total amount of macromolecules ) that is almost proportional to the total cell number , under a condition with a given finite resource , for comparison with experimental data in the batch culture condition ( fig.[fig : fig1].(c and d ) ) . in the numerical simulation , the condition with a given
, finite amount of substrates corresponding to the increase of cell number is implemented by introducing the dynamics of external substrate concentration to the original model . here
, @xmath28 is decreased as the substrates are replaced by the biomass , resulting in cell growth ( details are given in the _ supplementary information _ ) . at the beginning of the simulation , the amount of biomass ( i.e. , cell number ) stays almost constant , and then gradually starts increases exponentially . after the phase of exponential growth , substrates are consumed , and the biomass increase stops . then , over a long time span , the biomass stays at a nearly constant value , until it begins to slowly decrease .
finally , the degradation dominates and the biomass ( cell number ) falls off dramatically .
+ these successive transitions in the growth of biomass ( fig.[fig : fig1](c and d ) ) from initially inactive to the active , inactive , and death phases corresponds to those among the lag , log , stationary , and death phases . as the initial condition was chosen as the inactive phase under a condition of rich substrate availability , most of the active proteins are arrested in a complex at this point .
therefore , at the initial stage , dissociation of the complex into active proteins and inhibitors progress , and biomass is barely synthesized , even though rich substrate is available .
after the cell escapes this waiting mode , catalytic reactions from active proteins progress , leading to an exponential increase in biomass .
subsequently , the external substrate is depleted , and cells experience another transition from the active to inactive phase . at this point , the biomass decreases only slowly owing to the remaining substrate and stability of the active protein - inhibitor complex . however , after the substrate is depleted and the active protein and inhibitor are dissociated from the complex , the biomass decreases at a much faster rate , ultimately entering the death phase .
+ in the active phase with exponential growth , the present model exhibits classical growth laws , namely * ( i ) * monod s growth law , * ( ii ) * pirt s law , and * ( iii ) * growth rate vs. ribosome fraction ( see _ supplementary information _ fig .
in this section , we uncover the quantitative relationships among the basic quantities characterizing the transition between the active and inactive phases ; i.e. , lag time , starvation time , and growth rates .
we demonstrate that the theoretical predictions agree well with experimentally observed relationships .
+ first , we compute the dependency of lag time @xmath33 on starvation time @xmath34 and the maximum specific growth rate @xmath35 . up to time @xmath36 , the model cell is set in a substrate - rich condition @xmath37 , and stays at a steady state with exponential growth .
then , the external substrate is depleted to @xmath38 instantaneously .
the cell is exposed to this starvation condition up to starvation time @xmath39 .
subsequently , the substrate concentration @xmath28 instantaneously returns to @xmath40 .
after the substrate level is recovered , it takes a certain length of time for a cell to return to its original growth rate ( fig .
s2 ) , which is the lag time @xmath41 ( following the standard definition of introduced by penfold and pirt@xcite ) . given this ,
the dependency of @xmath41 on the starvation time @xmath42 and @xmath43 can be computed .
+ we found that @xmath41 increases proportionately to @xmath45 , as shown in fig.[fig : fig2](a ) .
the experimentally observed relationship between @xmath41 and @xmath42 is also plotted for comparison in fig.[fig : fig2](b ) , using reported data @xcite , which also exhibited @xmath46 dependency .
although this empirical dependency has been previously discussed , its theoretical origin has thus far not been uncovered .
+ indeed , the origin of @xmath47 is explained by noting the anomalous relaxation of inhibitor concentration , which is caused by the interaction between the active protein and inhibitor .
the sketch of this explanation is given below , and the analytic derivation is given in the _
supplementary information_. + first , consider the time course of chemical concentrations during starvation . in this condition ,
cell growth is inhibited by two factors : substrate depletion and deactivation of catalytic activity of the active protein . following the decrease in uptake due to depletion of @xmath28 , the concentration of @xmath0 decreases , resulting in a change in the balance between @xmath3 and @xmath1 .
( hereafter we adopt the notation such that @xmath3 , @xmath1 , and @xmath2 also denote the concentrations of corresponding chemicals ) . under the @xmath48 condition , the ratio of the synthesis of @xmath1 to @xmath3 increases . with an increase in @xmath1
, @xmath3 decreases due to the formation of a complex with @xmath1 . over time , more @xmath3 gets arrested , and the level of inactivation increases with the duration of starvation .
+ in this scenario , the increase of concentration @xmath1 is slow . considering that the complex formation reaction @xmath16 rapidly approaches its equilibrium , i.e. , @xmath49
, @xmath3 is roughly proportional to the inverse of @xmath1 ( recall @xmath26 ) , if @xmath1 is sufficiently large .
accordingly , the synthesis rate of @xmath1 , given by @xmath50 , is inversely proportional to its amount , i.e. , @xmath51 and thus @xmath52 hence , the inhibitor accumulation progresses with @xmath53 .
( note that due to @xmath0 depletion , the dilution effect is negligible . ) + next , we consider the time course for the resurrection after recovery of the external substrate . during resurrection , @xmath3 is increased while @xmath1 is reduced . since @xmath3 is strongly deactivated after starvation , the dilution effect from cell growth
is the only factor contributing to the reduction of @xmath1 .
noting @xmath54 and @xmath55 , the dilution effect is given by @xmath56 at the early stage of resurrection .
thus , the resurrection time course of @xmath1 is determined by the dynamics @xmath57 leading to the linear decrease of @xmath1 , i.e. , @xmath58 .
+ let us briefly recapitulate the argument presented so far .
the accumulated amount of component @xmath1 is proportional to @xmath45 , while during resurrection , the dilution of @xmath1 progresses linearly with time , which is required for the dissociation of @xmath3 and @xmath1 , leading to growth recovery . by combining these two estimates ,
the lag time satisfies @xmath46 .
second , the relationship @xmath60 is obtained by numerical simulation of our model , in line with experimental results @xcite ( fig.[fig : fig2](c and d ) ) .
+ this relationship @xmath61 is also explained by the characteristics of the resurrection time course .
the dilution rate of @xmath1 over time is given by @xmath62 , as mentioned above ; thus , at the early stage , @xmath63 .
in the substrate - rich condition , the substrate abundances are assumed to be saturated , so that @xmath64 holds because @xmath65 is satisfied .
thus , it follows that @xmath60 .
+ we also obtained an analytic estimation of the lag time as @xmath66 where @xmath67 ( see _
supplementary information _ for conditions and calculation ) . in this form
, the two relationships @xmath68 and @xmath69 are integrated .
+ the present theory also explains other experimental observations .
first , in predictive microbiology@xcite , the lag time to return the log phase from the stationary phase is regarded as the time span required to consume the @xmath70 accumulated during the stationary phase with the rate @xmath43 .
thus , the amount of @xmath70 is defined as the product of @xmath41 and @xmath43 . in our results ,
accumulated inhibitor @xmath1 needs to be consumed during the lag time @xmath41 , so that @xmath70 is interpreted as @xmath1 , whose time course agrees well with that of @xmath70 obtained experimentally ( see supplementary figure fig.s4 ) .
second , the tradeoff between the growth rate and tolerance for the starvation , experimentally observed @xcite is also derived from our theory ( see _ supplementary information _ ) .
so far , we have considered the dependence of lag time on the starvation time . however
, in addition to the starvation period , the starvation process itself , i.e. , the speed required to reduce the external substrate , has an influence on the lag time .
+ for this investigation , instead of the instantaneous depletion of the external substrate , its concentration is instead gradually decreased over time in a linear manner over the span @xmath71 , in contrast to the previous simulation procedure , which corresponds to @xmath72 .
then , the cell is placed under the substrate - poor condition for the duration @xmath42 , before the substrate is recovered , and the lag time @xmath41 is computed .
+ the dependence of the lag time @xmath41 on @xmath42 and @xmath71 is shown in fig.[fig : fig3](a ) .
while @xmath41 monotonically increases against @xmath42 for a given @xmath71 , it shows drastic dependence on @xmath71 .
if the external concentration of the substrate is reduced quickly ( i.e. , small @xmath71 ) , the lag time is rather small .
however , if the decrease in the external substrate concentration is slow ( i.e. , large @xmath71 ) , the lag time is much longer .
in addition , this transition from a short to long lag time is quite steep .
+ this transition against the timescale of the environmental change manifests itself in the time course of chemical concentrations ( see fig.[fig : fig3](b ) ) . with rapid environmental change
, @xmath0 decreases first , whereas with slow environmental change , @xmath3 decreases first .
in addition , the value of @xmath1 is quite different between the two cases , indicating that the speed of environmental change affects the degree of inhibition , i.e. , the extent to which active proteins are arrested by inhibitors to form a complex . + now , we provide an intuitive explanation for two distinct inhibition processes .
when @xmath28 starts to decrease , a cell is in the active phase in which @xmath3 is abundant .
if the environment changes sufficiently quickly , there is not enough time to synthesize the chemicals @xmath3 or @xmath1 , because of the lack of @xmath0 , and the concentrations of chemical species are frozen near the initial state with abundant @xmath3 . however ,
if the environmental change is slower than the rate of the chemical reaction , the concentration of the inhibitor @xmath1 ( active protein @xmath3 ) increases ( decreases ) , respectively .
hence , @xmath3 remains rich in the case of fast environmental change , whereas @xmath1 is rich for a slow environmental change . in the former case , when the substrate is increased again , the active proteins are ready to work , so that the lag time is short , which can be interpreted as a kind of `` freeze dry '' process .
note that the difference in chemical concentration caused by different @xmath71 is maintained for log time because in slow ( fast ) environmental change , chemical reactions are almost halted due to the decrease of @xmath3 ( @xmath0 ) , respectively .
thus , the difference of lag time remains even for large @xmath42 as fig.[fig : fig3](a ) ( the mechanism of this slow process is discussed in _ supplementary information_. ) .
+ this lag time difference can also be explained from the perspective of dynamical systems@xcite . for a given @xmath0 , the temporal evolution of @xmath3 and @xmath1
is given by the flow in the state space of @xmath73 .
examples of the flow are given in fig.[fig : fig4 ] .
the flow depicts @xmath74 , which determines the temporal evolution .
the flow is characterized by @xmath75 and @xmath76 nullclines , which are given by the curves satisfying @xmath77 and @xmath78 , as plotted in fig.[fig : fig4 ]
. + note that at a nullcline , the temporal change of one state variable ( either @xmath3 or @xmath1 ) vanishes .
thus , if two nullclines approach each other , then the time evolution of both the concentrations @xmath3 and @xmath1 are slowed down , and the point where two nullclines intersect corresponds to the steady state . as shown in fig .
[ fig : fig4 ] , nullclines come close together under the substrate - depleting condition , which gives a dynamical systems account of the slow process in the inactive phase discussed so far .
+ for a fast change ( i.e. , small @xmath71 , fig.[fig : fig4](a ) ) , @xmath0 is quickly reduced at the point where the two nullclines come close together .
then , the dynamics of @xmath73 follow the flow as shown in the figure .
first , @xmath1 decreases to reach the @xmath1-nullcline .
then , the state changes along the almost coalesced nullclines when the dynamics are slowed down .
thus , it takes a long time to decrease the @xmath3 concentration , so that at resumption of the substrate , sufficient @xmath3 can be utilized .
+ in contrast , for a slow change ( i.e. , large @xmath71 ) , the flow in @xmath73 gradually changes as shown in fig.[fig : fig4](b - d ) .
initially , the state @xmath73 stays at the substrate - rich steady state ( fig.[fig : fig4](b ) ) . due to the change in substrate concentration , two nullclines moderately
move and interchange their vertical locations .
since the movement of nullclines is slow , the decrease in @xmath3 progresses before the two nullclines come close together ( i.e. , before the process is slowed down ) .
the temporal evolution of @xmath3 and @xmath1 is slowed down only after this decrease in @xmath3(fig.[fig : fig4](c and d ) ) .
hence , the difference between the cases with small and large @xmath71 is determined by whether the nullclines almost coalesce before or after the @xmath3 decrease , respectively .
so far , we have considered the average change of chemical concentrations using the rate equation of chemical reactions .
however , the biochemical reaction is inherently stochastic , and thus the lag time is accordingly distributed .
this distribution was computed by carrying out a stochastic simulation of chemical kinetics using the gillespie algorithm@xcite .
+ we found that the distribution of lag time @xmath41 has a standard gaussian form for the shorter lag - time side but has an @xmath79 tail for the longer side ( fig.[fig : fig5 ] ) .
this exponential tail was also observed in experiments , as overlaid in fig.[fig : fig5 ] , which is adapted from reisman _
_ @xcite . in the present model ,
once the number of active proteins becomes small , more time is needed to recover the growth , so that the distribution of initial active protein abundances is expanded to a long - tailed distribution .
the agreement of the model with experimental data is relatively good for a short starvation time ( 24 hours and 48 hours ) but for longer times , the experimental data may suggest the existence of a much longer tail . +
we developed a coarse - grained model consisting of a substrate , autocatalytic active protein , inhibitor of the active protein , and active protein - inhibitor complex . in the steady state ,
the model shows distinct phases , i.e. , active , inactive , and death phases .
in addition , the temporal evolution of total biomass shows bacterial growth curve - like behavior .
the present model is not only consistent with the already - known growth laws in the active phase but also demonstrates two relationships , @xmath68 and @xmath60 , concerning the duration of the lag time @xmath41 .
although these two relationships have also been observed experimentally , their origins and underlying mechanisms had not yet been elucidated .
the present model can explain these relationships based on the formation of a complex between the active protein and inhibitor , whose increase in the starvation condition hinders the catalytic reaction .
the inactive phase , which corresponds to the stationary phase , as well as the above two laws are generally derived as long as the ratio of the synthesis of the inhibitor to that of the active protein is increased along with a decrease in the external substrate concentration .
this condition is also derived if the inhibitor is interpreted as a product of erroneous protein synthesis , where a proofreading mechanism to correct the error needing energy works inefficiently in a substrate - poor condition . + although the cell state with exponential growth has been extensively analyzed in previous theoretical models , the transition to the phase with suppressed growth has thus far not been theoretically explained .
our model , albeit simple , provides an essential mechanism for this transition as complex formation of active and inhibitor proteins , which can be experimentally tested .
+ moreover , the model predicts that the lag time differs depending on the rate of external depletion of the substrate , which can also be examined experimentally .
recently , the bimodal distribution of growth resumption time from the stationary phase was reported in a batch culture experiment@xcite .
the heterogeneous depletion of a substrate due to the spatial structure of a bacterial colony is thought to be a potent cause of this bimodality , while understanding of this concept is fairly underway .
since the present model shows different lag times for different rates of environmental change , it can provide a possible scenario for explaining this bimodality .
the authors would like to thank s.krishna , s.semsey , n.mitarai , a.kamimura , n.saito , and t. s. hatakeyama for useful discussions ; and i. l. reisman , n. balaban , and j.c .
augustin for providing data .
this research is partially supported by the platform for dynamic approaches to living system from japan agency for medical research and development ( amed ) , grant - in - aid for scientific research ( s ) ( 15h05746 from jsps ) , and the japan society for the promotion of science(16j10031 ) . | quantitative characterization of bacterial growth has gathered substantial attention since monod s pioneering study .
theoretical and experimental work has uncovered several laws for describing the log growth phase , in which the number of cells grows exponentially .
however , microorganism growth also exhibits lag , stationary , and death phases under starvation conditions , in which cell growth is highly suppressed , while quantitative laws or theories for such phases are underdeveloped .
in fact , models commonly adopted for the log phase that consist of autocatalytic chemical components , including ribosomes , can only show exponential growth or decay in a population , and phases that halt growth are not realized . here
, we propose a simple , coarse - grained cell model that includes inhibitor molecule species in addition to the autocatalytic active protein .
the inhibitor forms a complex with active proteins to suppress the catalytic process . depending on the nutrient condition
, the model exhibits the typical transition among the lag , log , stationary , and death phases .
furthermore , the lag time needed for growth recovery after starvation follows the square root of the starvation time and is inverse to the maximal growth rate , in agreement with experimental observations . moreover
, the distribution of lag time among cells shows an exponential tail , also consistent with experiments .
our theory further predicts strong dependence of lag time upon the speed of substrate depletion , which should be examined experimentally .
the present model and theoretical analysis provide universal growth laws beyond the log phase , offering insight into how cells halt growth without entering the death phase .
quantitative characterization of a cellular state , in terms of cellular growth rate , concentration of external resources , as well as abundances of specific components , has long been one of the major topics in cell biology , ever since the pioneering study by monod @xcite .
quantitative growth laws have been uncovered mainly by focusing on the microbial log phase in which the number of cells grows exponentially , including pirt s equation for yield and growth @xcite and the relationship between the fraction of ribosomal abundance and growth rate ( experimentally demonstrated by schaechter _ et al._@xcite , and theoretically rationalized by scott _
et al . _
@xcite ) , among others @xcite , in which the constraint to maintain steady growth leads to general relationships@xcite . in spite of the importance of the discovery of these universal laws , cells under poor conditions
exhibit different growth phases in which such relationships are violated .
indeed , in addition to the death phase , cells undergo a stationary phase under conditions of resource limitation , in which growth is drastically suppressed .
once cells enter the stationary phase , a certain time span is generally required to recover growth after resources are supplied , which is known as the lag phase .
although several quantities have been measured to characterize these phases , such as the length of lag time for resurrection , and the tolerance time for starvation or antibiotics @xcite , there has been no theory put forward to characterize the phase changes , and no corresponding quantitative laws have been established .
+ to develop a theory for bacterial physiology beyond the log phase , we first constructed a simple mathematical model that exhibits the changes among the lag , log , stationary , and death phases .
we then uncovered the quantitative characteristics of each of these phases in line with experimental observations . including bacterial growth curve , quantitative relationships of lag - time with starvation time and the maximal growth rate , exponentially - tailed distribution of lag - time , and trade - off between the growth rate and tolerance for the starvation .
these are formulated by the changes in inhibitor ( or mistranslated proteins ) chemicals in addition to changes in ribosomal proteins ( ribosomes ) .
the proposed model also allowed us to reach several experimentally testable predictions , including the dependence of lag time on the speed of the starvation process . |
the roots of optimal transport as a mathematical field go back to monge @xcite and kantorovich @xcite who established its modern formulation .
more recently it experienced a steep development prompted by brenier s theorem @xcite and the milestone phd - thesis of mccann @xcite .
the field is now famous for its striking applications in areas ranging from mathematical physics and pde - theory to geometric and functional inequalities .
we refer to @xcite for recent accounts of the theory .
very recently there has also been interest in optimal transport problems where the transport plan must satisfy additional martingale constraints .
such problems arise naturally in robust finance , but are also of independent mathematical interest , for example mirroring classical optimal transport they have important consequences for the study of martingale inequalities ( see e.g. @xcite ) .
early papers to investigate such problems include @xcite , and this topic is commonly referred to as martingale optimal transport . in mathematical finance , transport techniques
complement the skorokhod embedding approach to model - independent / robust finance ( we refer to @xcite for an overview , see @xcite for a link with optimal transport theory ) . in view of the central role taken by brenier s theorem in optimal transport ,
it is an intriguing question to obtain an analogous result also in the martingale setup . in this direction , @xcite have proposed a martingale version of brenier s monotone transport mapping . under certain structural properties of the underlying probability distributions ,
it was established in @xcite that the monotone martingale transport plan solves the given variational problem for all coupling functions satisfying the _ spence - mirrlees _ condition . in the present paper
we show that this is in fact true for general continuous distributions @xmath0 .
a basic fundamental result is that the monotone martingale transport plan is _
unique_. however the original derivation ( @xcite ) is intricate and relies on subtle properties of potential functions and a delicate approximation procedure .
we provide a short conceptual proof of this result .
this is based on a new connection with the classical skorokhod embedding problem which may be of independent interest .
on the canonical space @xmath1 , we denote by @xmath2 the canonical process , i.e. @xmath3 and @xmath4 for all @xmath5 .
we also denote by @xmath6 and @xmath7 the collection of all probability measures on @xmath8 and @xmath9 , respectively . for fixed @xmath10 with finite first moments ,
our interest is on the following subsets of @xmath7 : ( , ) & : = & \{ _ : x~_ , y~_ } , + ( , ) & : = & \{ ( , ) : ^[y|x]=x , -a.s . } .
the set @xmath11 is non - empty as it contains the product measure @xmath12 . by a classical result of strassen @xcite
, we also know that @xmath13 is non - empty if and only if @xmath14 in convex order , i.e. [ convexorder ] ( g)(g ) & & g. throughout we assume that @xmath15 is a measurable coupling function with @xmath16 for some @xmath17 and @xmath18 . here
@xmath19 for all @xmath5 . then @xmath20 $ ] is a well - defined scalar in @xmath21 .
the martingale optimal transport problem , as introduced in @xcite in the present discrete - time case and @xcite , is defined by : ( c ) : = & : = & _ ( , ) ^[c(x , y ) ] .
this problem is motivated by the problem of model - free superhedging in financial mathematics : @xmath22 where , denoting @xmath23 for all @xmath5 , @xmath24 the following result was established in @xcite .
[ thm : dualitymart ] let @xmath25 , and assume @xmath26 with @xmath16 for some lower semicontinuous functions @xmath17 , @xmath18
. then @xmath27 , and @xmath28 $ ] for some @xmath29 .
we also recall the variational result of @xcite which provides a characterization of the optimal martingale measure @xmath30 .
[ thm : variational ] let @xmath25 .
then , if @xmath29 is a solution of @xmath31 , there exists a support ( i.e. a borel set @xmath32 with @xmath33=1 $ ] ) such that for all @xmath34 , @xmath35 with finite support , and @xmath36 , we have ^^0 ^ & & (_0,_0 ) ^[y|x]=^^0[y|x ] . the fact that the absence of a duality gap ( as in theorem [ thm : dualitymart ] ) implies a variational result similar to the one given in theorem [ thm : variational ] is well known in the transport literature , see e.g. villani s book ( * ? ? ?
* p88 ) . extensions of theorem have been provided in @xcite , see @xcite for a variant applicable to skorokhod problem
. in particular zaev @xcite obtains ( as a special case of his results ) a version of theorem [ thm : variational ] with a simple proof under the assumption that the duality @xmath27 holds . for the convenience of the reader
we report the argument might fail ( e.g. if @xmath37 is not upper bounded in the sense of theorem [ thm : dualitymart ] ) and we state theorem [ thm : variational ] in this slightly more general form . ] from zaev @xcite .
pick sequences of admissible dual functions @xmath38 such that @xmath39 and fix @xmath40 such that @xmath20=\mathbf{p}.$ ] since @xmath41 and @xmath42 = \mu(\phi_n)+\nu(\psi_n ) \to \mathbf{p}= \e^{\p^*}[c(x , y)]\ ] ] it follows that @xmath43 tends to @xmath37 in @xmath44 .
passing to a subsequence which we denote again by @xmath45 we find that this convergence holds pointwise on a set @xmath46 with @xmath47 .
assume that @xmath34 , @xmath35 with finite support , and @xmath36 , and that @xmath48 satisfies @xmath49=\e^{\p^0}[y|x].$ ] note that then @xmath50=\e^{\p^0}[h(x)(y - x)]$ ] holds for an arbitrary function @xmath51 .
we thus obtain @xmath52 & = \lim_n \e^{\p^0}[\phi_n(x)+\psi_n(y)+h_n(x)(y - x ) ] \\ & = \lim_n \e^{\p}[\phi_n(x)+\psi_n(y)+h_n(x)(y - x ) ] \geq \e^{\p}[c(x , y ) ] , \end{aligned}\ ] ] hence @xmath46 is as required .
the following definition stems from @xcite .
[ def : monotone ] we say that @xmath53 is left - monotone ( resp . right - monotone ) if there exists a support ( i.e. a borel set @xmath54 with @xmath55=1 $ ] ) such that for all @xmath56 with @xmath57 ( resp .
@xmath58 ) , it must hold that @xmath59 .
the relevance of this notion is mainly due to the following extremality result which states that monotone martingale coupling measures are optimal for a class of martingale transport problems .
[ msm ] we say that a function @xmath60 satisfies the martingale spence - mirrlees condition if @xmath37 is measurable , continuously differentiable with respect to @xmath61 , and @xmath62 is strictly convex on @xmath8 for @xmath63 .
the martingale spence - mirrless condition was introduced in ( * ? ? ?
* remark 3.15 ) , cf .
also ( * ? ? ?
* definition 4.5 ) .
[ prop : optimal==>monotone ] assume that the performance function @xmath60 satisfies the martingale spence - mirrlees condition .
then , any solution @xmath30 of the martingale transport problem @xmath31 is left - monotone .
in particular , for all @xmath25 , there exists a left - monotone transference plan .
the last existence result is a consequence of the existence of a maximizer @xmath30 of the problem @xmath64 for some performance function @xmath37 satisfying the conditions of theorem [ thm : dualitymart ] .
let @xmath30 be a solution of the martingale transport problem @xmath31 , and suppose to the contrary that @xmath30 is not left - monotone .
let @xmath46 be an arbitrary support of @xmath30 . by definition of the notion of left - monotonicity
, we may find scalars @xmath57 and @xmath65 such that @xmath66 .
we then introduce : ^0 : = 12 + 12_(x,y ) , + : = 12 _ ( x , y ) + 12 , where @xmath67 is defined by y_0 + ( 1- ) y_1 & = & y. clearly , supp@xmath68 .
moreover , @xmath69 .
similarly , @xmath70 .
this shows that @xmath71 .
we also directly compute that @xmath72=\e^{\p}[y|x = x']=\lambda y_0+(1-\lambda)y_1 ~\mbox{and}~ \e^{\p^0}[y|x = x']=\e^{\p}[y|x = x]=y',\ ] ] which in view of the definition of @xmath73 , shows that @xmath74=\e^{\p}[y|x]$ ] .
we may then apply the variational theorem [ thm : variational ] , and conclude that @xmath75\ge\e^{\p}[c(x , y)]$ ] , i.e. g(x):=c(x , y_0)+(1-)c(x , y_1)-c(x , y ) & & g(x ) .
we now show that this inequality is in contradiction with the martingale spence - mirrlees condition .
indeed , as @xmath37 is continuously differentiable in @xmath61 , we have g(x)-g(x ) & = & _ x^x d > 0 , where the strict inequality follows from the strict convexity of the density @xmath76 in @xmath77 . for an atom - less measure @xmath78 , the following easy consequence , reported from @xcite ,
shows that left ( and right ) monotone martingale transport plans have a very simple structure .
namely that the support is concentrated on two graphs .
[ prop : leftmonotone2graphs ] let @xmath14 in convex order , and assume @xmath78 has no atoms .
let @xmath53 be a left - monotone transference plan . then there exist functions @xmath79 with @xmath80 , @xmath81 , and such that @xmath82(dy ) ~\mbox{with}~ q(x):=\frac{x - t_d(x)}{t_u(x)-t_d(x)}\1_{\{t_d(x)<t_u(x)\}}.\ ] ] moreover , the pair @xmath83 is unique @xmath84 and satisfies : 1 .
[ tu : increasing ] @xmath85 is non decreasing .
[ curtain ] if @xmath86 then @xmath87 . given the structure of left monotone transport plans for an atom - less probability measure @xmath78 , we call such a martingale transport measure a _ left monotone transport map_. proposition [ prop : leftmonotone2graphs ]
should be compared to the situation in standard optimal transport problem , where the frchet - hoeffding coupling defines a transport plan concentrated on a graph which solves simultaneously the monge and the kantorovitch problem .
clearly , one can not expect that martingale transport maps be concentrated on one single graph , as this would imply that the martingale is deterministic , and therefore constant which can happen only in the degenerate case @xmath88 .
the following uniqueness result for the martingale transport problem @xmath31 is a direct consequence of proposition [ prop : leftmonotone2graphs ] .
[ prop : uniquep ] let @xmath10 be such that @xmath14 in the convex order , and @xmath78 without atoms .
let @xmath60 be a performance function satisfying the martingale spence - mirrlees condition .
then , there exists at most one solution to the martingale transport problem @xmath64 .
suppose that @xmath64 has two solutions @xmath89 and @xmath90 in @xmath13 .
by thorem [ prop : optimal==>monotone ] and proposition [ prop : leftmonotone2graphs ] , it follows that both @xmath89 and @xmath90 are left monotone and concentrated on two graphs .
we next consider the probability measure @xmath91 .
clearly , @xmath92 , and @xmath93=\e^{\p_1}[c(x , y)]=\e^{\p_2}[c(x , y)]$ ] so that @xmath94 is also a solution of @xmath64 .
however , this is in contradiction with theorem [ prop : optimal==>monotone ] and proposition [ prop : leftmonotone2graphs ] , as @xmath94 is concentrated on more than two graphs .
it was established in @xcite that for fixed marginals @xmath0 there exists a _
unique _ left - monotone transference map in @xmath13 .
as the original argument is rather lengthy and maybe not entirely transparent , it seems worthwhile to revisit this basic ( but important ) result . in the rest of this paper , we assume that @xmath78 has no atoms and we fix @xmath95 and the corresponding @xmath96 as in proposition [ prop : leftmonotone2graphs ] . our aim is to prove that @xmath95 is the unique left - monotone transference plan in @xmath13 . to this end
, we use that a left - monotone martingale transport plan gives rise to a particular solution of the skorokhod embedding problem : for measures @xmath78 , @xmath97 in convex order the skorokhod problem is to construct a stopping time @xmath98 such for a brownian motion started in @xmath99 , the distribution of @xmath100 equals @xmath97 .
we refer to @xcite for recent surveys on the skorokhod embedding problem .
typically one is interested to find a minimal solution of the skorokhod problem , i.e. a stopping time which in addition to @xmath101 satisfies that for all @xmath102 with @xmath103 one has @xmath104 .
this notion was introduced by monroe @xcite in the case @xmath105 , and further extended to the case of a general starting law by cox @xcite . the fact that @xmath109 is clear by construction .
next , since @xmath110 is a hitting time , we now show that it is a minimal embedding of @xmath97 starting from @xmath78 .
minimality of hitting times with starting measure @xmath111 was observed by monroe @xcite ( just after his definition 1 ) , we report the full argument here in order to emphasize that this result extends trivially to an arbitrary starting measure @xmath78 . denote @xmath112 . for a stopping time @xmath113
, we have @xmath114 .
then , if @xmath115 also embeds @xmath97 , it follows that @xmath116 , and equivalently @xmath117 .
since @xmath110 is the first exit time from @xmath118 , this implies that , on the event set @xmath119 , we have @xmath120 while @xmath121 .
then @xmath117 implies that @xmath122 a.s . in the context of the skorokhod embedding problem a _
barrier _ is a measurable subset of @xmath123 such that for any point @xmath124 contained in the barrier , the whole line @xmath125 is a subset of the barrier .
we next introduce another stopping time defined as the first hitting time of this barrier : [ lctime ] _ r & : = & \{t : ( b_t - b_0 , b_t)r}. it is clear that @xmath127 .
the following crucial result shows that equality in fact holds .
we only focus on the non - trivial inequality @xmath130 . to see this
, we shall verify that the brownian path started in @xmath61 hits @xmath131 either in @xmath132 or in @xmath133 .
indeed , assume that a brownian path hits a different line of the barrier @xmath134 or @xmath135 , for some @xmath136 . from our construction ( see figure [ fig : rootsolution ] ) , we observe that , necessarily @xmath137 , and we have thus obtained an interpretation of the left - monotone transport plan in terms of a barrier - type solution to the skorokhod problem .
this interpretation is useful for our purpose since it allows us to use a short argument of loynes @xcite ( which in turn builds on root @xcite ) to show that there is only one left - monotone transference plan
. let @xmath89 , @xmath90 be left - monotone martingale transport plans in @xmath13 , with corresponding maps @xmath143 satisfying the conditions of proposition [ prop : leftmonotone2graphs ] , and denote by @xmath144 the corresponding barriers defined as in . then @xmath145 , a.s . for a set @xmath146 , we abbreviate @xmath147 , @xmath148
. denote k:= \{x : m_1(x ) < m_2(x ) } & & m_i(x):=\{m : ( m , x ) r^_i } , i=1,2 . fix a trajectory @xmath149 such that @xmath150
. then @xmath151 hits @xmath152 before it enters @xmath153 .
but then @xmath154 also hits @xmath155 before it enters @xmath156 .
hence @xmath157 as both stopping times embed the same measure , this implication is an equivalence almost surely , and we may set @xmath158 .
on @xmath159 we have @xmath160 while @xmath161 on @xmath162 .
then , for all borel subset @xmath146 : & = & + + & = & + + & = & , since @xmath163 . hence @xmath164 embeds @xmath97 .
similarly , we see that @xmath165 also embeds @xmath97 . since @xmath166 and @xmath167 are both minimal embeddings , we deduce that @xmath168 and thus @xmath169 . in @xcite , theorem [ unique ]
is proved without the assumption that @xmath78 has no atoms .
it is possible to use the present approach to establish also this more general result ; the basic idea is to represent the measure @xmath78 as an atom - less measure @xmath171 on the set @xmath172.\ ] ] since the extension of the above to arguments to the more general case is straightforward and the result is known from @xcite we do not elaborate . by proposition [ prop : min ] and lemma [ lem :
taur = taup ] , @xmath142 is a minimal embedding of @xmath97 with starting measure @xmath78 . since @xmath174=\e[b_0]$ ] , we know from lemma 12 and theorem 17 in cox @xcite that minimality is equivalent to the uniform integrability of the process @xmath175 . for the convenience of the reader , we also provide a direct justification of the uniform integrability in our setting .
observe that @xmath176 $ ] , a.s .
then , conditional on @xmath177 , the process @xmath175 is a bounded martingale . by the jensen inequality
, this provides , . in particular
, it follows that for any constant @xmath178 : & & , which provides the uniform integrability of the process @xmath175 .
a. m. g. cox . extending chacon - walsh
: minimality and generalised starting distributions . in _
sminaire de probabilits xli _ , volume 1934 of _ lecture notes in math .
_ , pages 233264 .
springer , berlin , 2008 .
d. hobson .
the skorokhod embedding problem and model - independent bounds for option prices . in _
paris - princeton lectures on mathematical finance 2010 _ , volume 2003 of _ lecture notes in math .
_ , pages 267318 .
springer , berlin , 2011 . | we show that the left - monotone martingale coupling is optimal for any given performance function satisfying the martingale version of the spence - mirrlees condition , without assuming additional structural conditions on the marginals .
we also give a new interpretation of the left monotone coupling in terms of skorokhod embedding which allows us to give a short proof of uniqueness .
_ keywords _ optimal transport ; martingales ; skorokhod embedding _ ams 2010 subject classification _ 60g42 ; 49n05 |
the thermodynamical property of spacetime is an important hint for the quantization of gravity . starting from hawking s discovery of black hole
s radiation @xcite , a theory of thermodynamics of spacetime is being constructed gradually .
recently , the second law of this thermodynamics was generalized to the covariant entropy bound conjecture @xcite .
it states that the entropy flux @xmath0 through any null hypersurface generated by geodesics with non - positive expansion , emanating orthogonally from a two - dimensional ( 2d ) spacelike surface of area @xmath1 , must satisfy @xmath2 where @xmath3 is the planck length .
here and in what follows , we adopt the units @xmath4 .
soon , flanagan , marolf and wald @xcite proposed a new version of the entropy bound conjecture .
if one allows the geodesics generating the null hypersurface from a 2d spacelike surface of area @xmath1 to terminate at another 2d spacelike surface of area @xmath5 before coming to a caustic , boundary or singularity of spacetime , one can replace the above conjecture with @xmath6 more recently , he and zhang related these conjectures to dynamical horizon and proposed a covariant entropy bound conjecture on the cosmological dynamical horizon @xcite : let @xmath7 be the area of the cosmological dynamical horizon at cosmological time @xmath8 , then the entropy flux @xmath0 through the cosmological dynamical horizon between time @xmath8 and @xmath9 ( @xmath10 ) must satisfy @xmath11 if the dominant energy condition holds for matter . as a non - perturbative and background - independent quantization of gravity , the loop quantum gravity ( lqg ) developed rapidly in recent years .
its cosmological version , the loop quantum cosmology ( lqc ) achieved many successes , including resolution of the classical singularity @xcite , quantum suppression of classical chaotic behavior near singularities @xcite , generic phase of inflation @xcite and more ( for example @xcite ) .
since it has been suggested that the holographic principle is a powerful hint and should be used as an essential building block for any quantum gravity theory @xcite , it is important and tempting to investigate the covariant entropy bound conjecture in the framework of the lqc , which is a successful application of the non - perturbative quantum gravity scheme the lqg .
the authors of @xcite investigated the bousso s covariant entropy bound @xcite with a cosmology filled with photon gas and found that the conjecture is violated near the big bang in the classical scenario .
but they found the lqc can protect this conjecture even in the deep quantum region . in @xcite , he and zhang proposed a new version of the entropy bound conjecture for the dynamical horizon in cosmology and validated it through a cosmology filled with adiabatic perfect fluid , governed by the classical einstein equation when the universe is far away from the big bang singularity .
but when the universe approaches the big bang singularity , the strong quantum fluctuation does break down their conjecture .
in analogy to ashtekar and wilson - ewing s result @xcite , one may wonder if he and zhang s conjecture can also be protected by the quantum geometry effect of the lqg . following @xcite , we use photon gas as an example to investigate this problem . as expected , we find that the loop quantum effects can indeed protect the conjecture .
besides the result of @xcite , our result presents one more evidence for the consistence between the loop gravity and the covariant entropy conjecture .
this paper is organized as follows . in sec .
[ sec.2 ] , we briefly review the framework of the effective lqc and describe the covariant entropy bound conjecture proposed by he and zhang @xcite . then in sec .
[ sec.3 ] , we test this conjecture with cosmology filled with photon gas , and show that the lqc is able to protect the conjecture in all .
we conclude the paper in sec .
[ sec.4 ] and discuss the implications .
to a good degree of approximation , the universe can be described by the well - known friedmann - robertson - walker ( frw ) metric , @xmath12 which describes a homogeneous and isotropic universe . here the spatial curvature @xmath13 correspond to open , flat and closed universes respectively , and @xmath14 is the scale factor of the universe . for simplicity
we focus on a flat universe in this paper . in the lqc ,
the phase space for a flat universe is spanned by the coordinates @xmath15 , being the gravitational gauge connection , and @xmath16 , being the densitized triad .
@xmath17 is the barbero - immirzi parameter @xcite .
then the effective hamiltonian in the lqc is given by @xcite @xmath18 the variable @xmath19 corresponds to the dimensionless length of the edge of the elementary loop and is given by @xmath20 where @xmath21 and @xmath22 depend on the particular scheme in the holonomy corrections . in this paper
we take the @xmath19-scheme , which gives @xmath23 and @xmath24 . with this effective hamiltonian
, we have the canonical equation @xmath25 or , @xmath26 combining with the constraint on the hamiltonian , @xmath27 , we obtain the modified friedmann equation , @xmath28 in terms of the hubble rate @xmath29 , where @xmath30 and @xmath31 .
@xmath32 is the critical energy density coming from the quantum effect in the lqc .
it is a large quantity and when it goes to infinity all the quantum effects disappear .
according to @xcite , the cosmological dynamical horizon @xcite is defined geometrically as a three - dimensional hypersurface foliated by spheres , where at least one orthogonal null congruence with vanishing expansion exists . for a sphere characterized by any value of @xmath33 , there are two future directed null directions @xmath34 satisfying geodesic equation @xmath35 .
the expansion of these null directions is @xmath36 where the dot denotes differential with respect to @xmath8 , and the sign @xmath37 represents the null direction pointing to larger ( smaller ) values of @xmath38 . for an expanding universe ,
i.e. @xmath39 , @xmath40 determines the location of the dynamical horizon , @xmath41 , by the definition of dynamical horizon given above .
the lqc replaces the big bang with the big bounce , so the universe is symmetric with respect to the point of the bounce , expanding on one side of the bounce and contracting on the other side .
the dynamical horizon in the contracting stage of the lqc corresponds to @xmath42 , and all of the relations are similar to the ones given here . in this paper
we only consider the expanding stage for the lqc , but note that the contracting stage is the same .
since the area of the dynamical horizon is @xmath43 , the covariant entropy bound conjecture in our question becomes @xmath44,\end{aligned}\ ] ] where @xmath0 is the entropy flux through the dynamical horizon between cosmological time @xmath8 and @xmath9 ( @xmath10 ) , and @xmath45 is the hubble parameter . considering that the cosmology model discussed here is isotropic and homogeneous
, we can write the entropy current vector as @xmath46 where @xmath47 is the ordinary comoving entropy density , independent of space .
if the entropy current of the perfect fluid is conserved , i.e. , @xmath48 , @xmath47 will be independent of @xmath8 as well .
for simplicity we restrict ourselves to this special case . the entropy flux through the dynamical horizon ( shown in fig.[fig1 ] )
is given by @xmath49 where @xmath50 is the spacetime volume 4-form .
so the conjecture is reduced to @xmath51 and @xmath9 respectively .
the dashed lines are the entropy current.,scaledwidth=45.0% ]
given that the frw universe is filled with photon gas , the energy momentum tensor can be expressed as @xmath52\right\ } .\end{aligned}\ ] ] the pressure @xmath53 and the energy density @xmath54 satisfy a fixed equation of state @xmath55 where the constant @xmath56 . from @xmath57 , we have the conservation equation @xmath58 the comoving entropy density @xmath47 is given by @xmath59 and @xmath54 depends only on the temperature @xmath60 , @xmath61 where @xmath62 is a dimensionless constant depending on the density of energy state of the perfect fluid . for photon gas @xmath63 .
plugging above thermodynamics relation into equation ( [ entropydensity ] ) we get @xmath64 . written the above conservation equation as @xmath65 we have an integration constant @xmath66 . then @xmath67 . combining our equation of state eq .
( [ eos ] ) with the above conservation equation , we get the relationship between @xmath54 and the hubble parameter , @xmath68 substituting the above relation ( [ relationh ] ) into the modified friedmann equation ( [ friemann ] ) , we can get @xmath69 where @xmath70 is an integration constant without direct physical significance , and we can always drop it by resetting the time coordinate . setting @xmath71
gives @xmath72 with the definition of the hubble parameter , we can integrate once again to get @xmath73^{\frac{1}{3(1+\omega)}}.\end{aligned}\ ] ] when @xmath32 goes to infinity , all of the above solutions become the same as the classical ones , while we use universe time @xmath8 in this paper .
@xmath74 can be negative which divides the discussion into two cases .
@xmath8 is always positive and makes the discussion simpler . ]
presented in @xcite . in the classical scenario , @xmath75 when @xmath76 , @xmath77 which is a decreasing function of @xmath8 , so the conjecture breaks down when the universe approaches the big bang singularity . and its derivative respect to @xmath78 for photon gas.,scaledwidth=45.0% ]
we introduce a new variable @xmath79 for the lqc to simplify the above expressions to @xmath80 then @xmath81 .
\nonumber \\\end{aligned}\ ] ] it is obvious that the necessary and sufficient condition for meeting the covariant entropy bound conjecture is that the above expression increases with @xmath78 . in order to investigate the monotone property of above function , we plot @xmath82 itself and its derivative respect to @xmath78 in fig.[photon ] .
the minimal value of the derivative is about @xmath83 .
the covariant entropy bound conjecture for dynamical horizon in cosmology is fully protected by loop quantum effect .
the covariant entropy bound conjecture comes from the holographic principle and is an important hint for the quantum gravity theory . in the recent years we have witnessed more and more success of the loop quantum gravity , especially for the problem of the big bang singularity in cosmology .
the entropy bound conjecture usually breaks down in the strong gravity region of spacetime where the quantum fluctuation is strong , and one would expect the loop quantum correction to protect the conjecture from the quantum fluctuation
. and ashtekar and wilson - ewing do find a result in @xcite which is consistent with above expectation . in this paper , we generalized the covariant entropy conjecture for the cosmological dynamical horizon proposed in @xcite to the loop quantum cosmology scenario .
we found that the quantum geometry effects of the loop quantum gravity can also protect the conjecture .
our result gives out one more evidence for the consistence of covariant entropy conjecture and loop quantum gravity theory .
this adds one more encouraging result of loop quantum gravity theory besides previous ones .
hua - hui xiong and jian - yang zhu , phys .
d * 75 * , 084023 ( 2007 ) .
li - fang li and jian - yang zhu , phys .
d * 79 * , 044011 ( 2009 ) ; phys .
d * 79 * , 124011 ( 2009 ) ; advances in high energy physics , doi:10.1155/2009/905705 . | the covariant entropy bound conjecture is an important hint for the quantum gravity , with several versions available in the literature .
for cosmology , ashtekar and wilson - ewing ever show the consistence between the loop gravity theory and one version of this conjecture .
recently , s. he and h. zhang proposed a version for the dynamical horizon of the universe , which validates the entropy bound conjecture for the cosmology filled with perfect fluid in the classical scenario when the universe is far away from the big bang singularity .
however , their conjecture breaks down near big bang region .
we examine this conjecture in the context of the loop quantum cosmology . with the example of photon gas
, this conjecture is protected by the quantum geometry effects as expected . |
the heavy fermion metal ybrh@xmath0si@xmath0 shows thermodynamic and transport properties which , at low temperatures , are determined by the hybridization of 4@xmath8 states and conduction electrons resulting in a kondo interaction with a single - ion kondo temperature @xmath9 k.@xcite besides its quite remarkable magnetic behaviors , like weak antiferromagnetic ( afm ) order below @xmath10 mk , a field induced afm quantum critical point at @xmath11 mt , and a highly enhanced sommerfeld - wilson ratio,@xcite the existence of a well - defined electron spin resonance ( esr ) signal at temperatures _ below _
@xmath12 with pronounced yb@xmath4 kondo - ion character triggered new theoretical studies of the dynamical properties of local and itinerant magnetism in kondo lattice systems.@xcite in these theories the origin of the resonance was intimately connected to the kondo interaction and strong ferromagnetic correlations .
the relevance of both effects were experimentally verified by esr results on lu - diluted ybrh@xmath0si@xmath0,@xcite esr experiments under pressure and upon co doping ( @xmath13),@xcite and the esr of the ferromagnetic kondo lattice cerupo.@xcite + recently the compound series yb(rh@xmath1co@xmath14)@xmath0si@xmath0 was investigated for the whole concentration range @xmath15 . whereas the crystal structure remains tetragonal ( space group i4/mmm ) the magnetic , thermal , and transport properties show clear variations with @xmath16:@xcite it was shown that the kondo interaction can largely be suppressed by substituting rh with the smaller isoelectronic co which , at the same time , enhances @xmath17 and the size of the ordered moment , thus stabilizing the afm order by suppressing the kondo screening.@xcite for ybco@xmath0si@xmath0 with @xmath18 k a stable trivalent yb state with well defined crystal electric field ( cef ) levels was evidenced by susceptibility @xcite , mbauer results @xcite , inelastic neutron - scattering experiments @xcite , specific heat measurements @xcite as well as by photoemission spectroscopy @xcite .
the co - ion itself remains non - magnetic because of a strong co - si hybridization.@xcite in the magnetically ordered state of ybco@xmath0si@xmath0 a complex magnetic phase diagram with significant basal - plane anisotropy is evidenced by the magnetoresistance @xcite and magnetization data @xcite .
powder neutron diffraction revealed a complex magnetic structure with two phases with different propagation vectors.@xcite a first - order phase transition from an incommensurable arrangement of moments below @xmath17 into a commensurable one below @xmath19 was identified.@xcite for ybrh@xmath0si@xmath0 the stronger 4@xmath8-conduction electron hybridization leads to a small deviation of the valency from the trivalent state .
photoemission spectroscopy results suggested a mean valency of @xmath20.@xcite yet , the magnetic structure of ybrh@xmath0si@xmath0 is not known due to its very small ordered moment .
some hints could be obtained from the knowledge of the magnetic structure in ybco@xmath0si@xmath0 since within an isoelectronic series of compounds magnetic structures are often closely related .
+ in this paper we first present the esr results of ybco@xmath0si@xmath0 and analyze the data in the context of further experimental results . we continue with the series yb(rh@xmath1co@xmath14)@xmath0si@xmath0 where we found a sensitive dependence of the esr parameters on the co -content .
for our esr measurements we used clean in - flux grown single crystals of yb(rh@xmath1co@xmath14)@xmath0si@xmath0 which all had their @xmath21-axis perpendicular to the main surface of the platelet - shaped crystals .
thermal and magnetic properties as well as electrical resistivity have been thoroughly investigated and were described elsewhere.@xcite + esr detects the absorbed power @xmath22 of a magnetic microwave field @xmath23 as a function of a transverse external static magnetic field @xmath24 . to improve the signal - to - noise ratio , a lock - in technique is used by modulating the static field , which yields the derivative of the resonance signal @xmath25 .
the esr measurements were performed with a standard continuous wave spectrometer at x - band frequencies ( @xmath26 ) by using a cylindrical resonator in te@xmath27 mode or , for temperatures down to 1.5 k , a split - ring resonator .
the temperature was varied between @xmath28 using @xmath29he - flow - type cryostats .
the esr measurements at q - band frequencies ( 34 ghz ) turned out to provide much less definite spectra .
their linewidths and @xmath6-factors are found to be comparable to the x - band results regarding temperature dependencies and absolute values . + all recorded esr spectra were analyzed by fitting them with a metallic lorentzian line shape.@xcite from these fits the following esr parameters were extracted : linewidth @xmath30 ( hwhm ) , the esr @xmath6-factor ( as given by the resonance field @xmath31 in the resonance condition @xmath32 ) , the asymmetry parameter @xmath33 ( describing the dispersive contribution in metallic samples ) , and the intensity ( spin susceptibility , given by the area @xmath34 under the esr absorption ) .
the esr resonance field displays for the whole series yb(rh@xmath1co@xmath2)@xmath3si@xmath3 ( @xmath35 ) a pronounced anisotropy upon rotating the crystalline @xmath7-axis from parallel ( @xmath36 ) towards perpendicular ( @xmath37 ) to the magnetic field .
this confirms the local character of the signal and verifies a typical resonance feature of yb - ions which are located on lattice sites with tetragonal point symmetry.@xcite fig.[fig1 ] shows typical esr spectra of ybco@xmath0si@xmath0 for selected directions of the static magnetic field @xmath24 ( angle @xmath38 ) and the microwave magnetic field @xmath23 ( angle @xmath39 ) .
all spectra could be well described by a metallic lorentzian lineshape ( solid line ) . within experimental accuracy linewidth and resonance field
remain unchanged if the @xmath7-axis is rotated from @xmath40 to @xmath41 .
interestingly , in that case , the esr intensity only varies by a factor of 1.5 .
this is not expected from the considerable uniaxial anisotropy shown in the inserted graph for @xmath42 : with the observed values @xmath43 and @xmath44 the intensity should vary by a factor of @xmath45 .
the origin of this discrepancy which is even more pronounced in ybir@xmath0si@xmath0 and ybrh@xmath0si@xmath0 @xcite remains unclear .
+ si@xmath0 at @xmath46 k for the field orientation @xmath47-axis ( @xmath37 and @xmath39-orientation of the basal plane to the microwave magnetic field ) and @xmath48-axis ( @xmath36 ) .
solid lines indicate lorentzian lines ( @xmath37 : same resonance field and linewidth for @xmath49 ) .
inset shows the @xmath38-dependence of the resonance - field - calculated @xmath6-factor with uniaxial behavior indicated by the solid line .
peak at 0.325 t arises from background signals of the resonator . ] and esr @xmath6-factor of ybco@xmath0si@xmath0 above the antiferromagnetic ordering temperature @xmath17 for the field orientation @xmath50-axis and @xmath51-axis .
dashed lines in @xmath52 shows the mean field expectation as deduced from @xmath53 . here , a larger crystal than the one of fig .
[ fig1 ] was used , hence yielding stronger lines and a higher data accuracy . ]
the observed @xmath6-values nicely agree with the values expected from the saturation magnetization @xcite : @xmath54 ( @xmath55 ) and @xmath56 ( @xmath57 ) .
these @xmath6-factors can be well accounted for by a cef model using a ground state doublet with a @xmath58 symmetry and a strong mixing of @xmath59 and @xmath60 wave functions.@xcite the deviations of the esr @xmath6-values from @xmath61 are partly due to the pronounced temperature dependence of the @xmath6-values as shown in fig .
[ fig2 ] . as a consequence of the anisotropic interactions between the yb - ions the internal magnetic field leads to a temperature dependent shift of the resonance fields
. this can reasonably be described by a molecular field model @xcite @xmath62 as shown by the dashed lines in fig .
[ fig2 ] . here , the weiss temperatures @xmath63 are obtained by a curie - weiss [ @xmath64 plot of the susceptibilities @xmath65 below @xmath66 k measured at the resonance magnetic fields of 0.46 t and 0.35 t for @xmath67 and @xmath68 , respectively : @xmath69 and @xmath70 . furthermore we used @xmath71 and @xmath72 .
although the dashed lines in fig . [ fig2 ] do not agree perfectly with the data they , however , reflect the correct tendency .
namely , towards low temperatures the @xmath6-values tend to merge each other which is a consequence of @xmath73 .
note , in ybir@xmath0si@xmath0 @xmath74 and thus we observed a mutual divergent behavior of @xmath75 and @xmath76.@xcite + the temperature dependence of the esr linewidth clearly differs at low temperatures from the linewidth behaviors of ybir@xmath0si@xmath0 and ybrh@xmath0si@xmath0 where a continuous decrease or a saturation towards low temperatures was observed @xcite .
as shown in fig .
[ fig2 ] the linewidth of ybco@xmath0si@xmath0 starts to increase below @xmath77 k indicating the effect of critical spin fluctuations due to magnetic order below @xmath78 k. above @xmath77 k the linewidth continuously increases , similar to ybir@xmath0si@xmath0 and ybrh@xmath0si@xmath0 . for @xmath79 k
the esr signal is too weak and too broad to be detected .
a comparison of x - band with q - band data of ybco@xmath0si@xmath0 shows no change in linewidth and @xmath6-value regarding absolute value and temperature dependence . and
solid lines depict metallic lorentzian line shapes . ]
figure [ fig3 ] presents the esr signals of the series yb(rh@xmath1co@xmath2)@xmath3si@xmath3 at @xmath81 k and for the magnetic field @xmath47-axis , i.e. the orientation at which the resonance field is the smallest .
all depicted spectra show well defined metallic lorentzian shapes .
the ratio of dispersion to absorption varies between 0.64 and 1 without clear relation to the @xmath16 values which may be explained by experimental reasons like different resonator filling factors .
+ -factor , linewidth @xmath30 and relaxation rate @xmath82 for the field orientations @xmath83-axis and for @xmath46 k. solid lines are guides for the eyes .
insert : evolution of @xmath6 factor anisotropy with co - concentration in comparison with the expected @xmath6 factors ( solid lines ) which correspond to @xmath84 and @xmath58 groundstate symmetries for a yb@xmath4 ion in a tetragonal crystal field . ] linewidth and resonance field show an overall continuous change with the co - content @xmath16 as shown in fig .
[ fig4 ] for @xmath46 k. the anisotropy of both quantities reduces with increasing @xmath16 .
note , that @xmath85 shows a narrowing with increasing co - doping , providing evidence that the relaxation due to disorder is not the dominant broadening mechanism .
the relaxation rate which is determined by @xmath82 shows no clear anisotropy but a remarkable increase with increasing @xmath16 up to @xmath86 and a slight decreasing tendency with further increasing @xmath16 up to 1 . both @xmath87 and @xmath88 evolve with @xmath16 with no pronounced anomalies , especially around @xmath89 where @xmath17 shows a minimum.@xcite the inset of fig .
[ fig4 ] provides a comparison of the measured @xmath6-values at @xmath46 k ( symbols ) with those expected for the two possible kramers doublets @xcite with irreducible representations @xmath84 or @xmath58 .
the experimental values are close to the theoretical values for either @xmath84 or @xmath58 doublets for all co - contents .
thus , the esr @xmath6-values can be well accounted for by a localized model .
however , these results do not allow a clear distinction between @xmath84 and @xmath58 .
the latter was established for the ground state of ybrh@xmath0si@xmath0 by angle - resolved photoemission investigations of the energy dispersion of the crystal - field - split 4@xmath8 states @xcite and by an analysis of the quadrupole moment @xcite .
furthermore , the analysis of all available data on ybco@xmath0si@xmath0 using a standard cef model indicate a @xmath58 ground state for ybco@xmath0si@xmath0 too.@xcite all these results indicate a smooth and continuous evolution of the crystal electric field from ybrh@xmath0si@xmath0 to ybco@xmath0si@xmath0 without change of the symmetry of the ground state doublet . the evolution of the linewidth and its temperature dependence is shown in fig . [ fig5 ] .
different characteristic behaviors suggest three temperature regimes : a linear temperature regime which separates a regime below @xmath90 k , for the field orientation @xmath50-axis .
dashed lines show for @xmath91 the maximum linear increase .
the solid line describes the low temperature part of the @xmath92 data with a @xmath93 behavior ( @xmath94 mt / k ) , see main text sec .
[ disc ] . ] characterized by tendencies to saturation or to low temperature broadening , from a regime above @xmath95 k where an additional linewidth broadening indicates the influence of relaxation via the first excited cef - level.@xcite the linear linewidth behavior of the middle temperature regime reminds to a korringa law of local - moment relaxation toward conduction electrons .
this is displayed in fig.[fig5 ] by the linear dashed line for the @xmath91 sample which among all samples displays the largest temperature slope @xmath96 of 22 mt / k .
the linearly extrapolated linewidth at zero - temperature has roughly the same value for the samples with @xmath97 . with further increasing @xmath16 @xmath98 decreases again and for @xmath99 is approximately the same as for @xmath100 . in the temperature regime below @xmath90
k a deviation from linearity is observed for @xmath101 @xcite and @xmath99 .
this is most obviously seen for the @xmath92 linewidth , showing a considerable broadening above the magnetic ordering temperature . for @xmath102 and @xmath103
small positive curvatures in the low temperature linewidth behaviors rather seem to indicate a tendency for saturation .
this tendency was discussed for @xmath104 in terms of a relation between the zero - temperature linewidth to the residual electrical resistivity.@xcite + the behavior of the esr @xmath6-factor is shown on a logarithmic temperature scale in fig .
[ gtemp ] . in general ,
as indicated by the dashed lines , all the @xmath6-data are characterized by a logarithmic downturn towards low temperatures .
-factor for the field orientation @xmath47-axis .
dashed lines guide the eyes and emphasize the trend in the low temperature behavior . ]
this is most nicely developed for @xmath105 where it was related to a small characteristic energy for the ground state kramers doublet.@xcite whereas in this low temperature region the temperature variation is largest for @xmath106 the absolute @xmath6-values steadily decrease with co doping ( see also fig . [ fig4 ] ) , indicating the effect of changes in the crystal electric field .
the influence of the first excited crystal - field levels may be seen in the decrease of the @xmath6-factors at higher temperatures , being quite sensitive to the co - content .
unfortunately , due to the insufficient experimental accuracy , further information on these excited levels could not be obtained .
+ the esr intensity per yb - ion does not change within experimental accuracy across the doping series .
previous detailed investigations of the number of esr active yb spins in ybrh@xmath0si@xmath0 have shown that _ all _ yb ions contribute to the esr signal.@xcite hence , for yb(rh@xmath107co@xmath108)@xmath3si@xmath3 , where any evidence for co magnetism is absent , the esr properties are always determined by _ all _ yb ions despite the large variation of the kondo interaction with co - doping.@xcite
collecting information on the competition of the local kondo and the intersite rkky exchange interactions is essential for understanding the magnetic behavior of yb(rh@xmath1co@xmath2)@xmath3si@xmath3 .
the doping with co enhances the size of the ordered moment and stabilizes the afm order by weakening the kondo interaction.@xcite the observed esr parameters show dependencies on the co - content which are sensitively related to the kondo effect , to the anisotropy induced by the crystal field as well as to the presence and character of magnetic phase transitions below the investigated temperature range . in the following
this will be discussed by analyzing the linewidth parameters shown in fig.[pd ] in terms of the relevant magnetic properties of yb(rh@xmath107co@xmath108)@xmath3si@xmath3 . at @xmath81 k and linewidth temperature slope @xmath98 between 4 and 8 k ( 9 and 13 k ; x=1 ) with the characteristic temperature @xmath109 deduced from the @xmath110- specific heat and the low - temperature , in - plane weiss temperature @xmath111 of yb(rh@xmath107co@xmath108)@xmath3si@xmath3 ( data from ref . ) .
bottom frame shows magnetic phase diagram @xcite as determined from anomalies of @xmath53 and @xmath112 . below temperatures @xmath17 and @xmath113 magnetic order appears to be incommensurate and commensurate , respectively .
while the transition at @xmath17 is always second - order , the transition at @xmath113 is second order for @xmath114 and first - order for @xmath115 . ]
we start with discussing the relation between the linewidth and the characteristic temperature @xmath109 deduced from the magnetic entropy which reflects the energy scale of all exchange interactions acting on the lowest cef doublet.@xcite we have chosen the linewidth @xmath116 at @xmath81 k as one possible measure of the spin fluctuations which dominate the linewidth at low temperatures . as shown in the upper frame of fig .
[ pd ] @xmath117 and @xmath118 show a similar behavior . for low co
- contents this similarity suggests that a decrease of the kondo energy scale @xmath12 leads to an increase of the esr linewidth because @xmath12 dominates @xmath109 for co - concentrations @xmath119 . in this respect
it is important to mention also the effect of anisotropy of the exchange interactions to the linewidth .
it has been shown that a collective spin mode of kondo ions and conduction electrons enables a narrow linewidth in a kondo lattice despite a strong magnetic anisotropy.@xcite hence , despite the anisotropy ( as seen in the ratio of @xmath120 , fig .
[ fig4 ] ) is largest at @xmath105 , there the linewidth reaches its smallest value , supporting the collective spin mode model .
note , that with increasing @xmath16 the increasing contribution of disorder to the linewidth should be considered as well .
a discrimination of kondo- and disorder contributions was qualitatively accomplished by a comparison of the linewidth - resistivity relation of pressurized ybrh@xmath0si@xmath0 and yb(rh@xmath107co@xmath108)@xmath3si@xmath3 with @xmath13.@xcite the result suggests that disorder is not the dominant contribution in @xmath117 .
+ at high co - contents the esr linewidth should be considered in terms of a strongly suppressed @xmath12 because @xmath109 gets dominated by rkky exchange interactions for @xmath121 .
a strongly reduced kondo interaction is established from photoemission spectroscopy : ybco@xmath0si@xmath0 ( with @xmath122 k @xcite ) shows a rather weak interaction between the 4@xmath8 and the valence states in comparison to ybrh@xmath0si@xmath0.@xcite we discuss the role of the rkky interactions for the linewidth in terms of ferromagnetic fluctuations which were shown to play an important role for the observability of a spin resonance in kondo lattice systems.@xcite the evolution of ferromagnetic fluctuations in yb(rh@xmath107co@xmath108)@xmath3si@xmath3 are reflected , for instance , in the weiss temperature @xmath111 characterizing the in - plane magnetic susceptibility at low temperatures .
@xmath123 shows a clear maximum around @xmath124 , where it even changes sign and becomes positive , indicating the dominance of ferromagnetic exchange interactions . as documented in the middle frame of fig .
[ pd ] this behavior is well correlated with the in - plane linear temperature slope of the linewidth , @xmath125 . for low @xmath16
we interpret this correlation , similar to @xmath117 , by a decreasing kondo interaction up to @xmath126 .
+ at this point it is worth to note that in compounds with no kondo lattice behavior ( very small kondo energy scales ) such as ybrh @xcite or ybco@xmath0si@xmath0 narrow esr lines emerge from ferromagnetic correlated 4@xmath8 spins whereas antiferromagnetic correlations as in yb@xmath127rh@xmath128ge@xmath129 @xcite lead to an esr linewidth too broad to be observable .
that means , for yb(rh@xmath107co@xmath108)@xmath3si@xmath3 above @xmath126 , the ferromagnetic correlations among the 4@xmath8 moments presumably start to control the linewidth , preventing @xmath125 to get further increased which may be expected from an increasing influence of the rkky interaction .
moreover , at high co - contents , narrow lines are also allowed by an anisotropy @xmath120 being much smaller than for low co - contents .
this is consistent with a constant or slightly decreasing ( for @xmath121 ) relaxation rate , see fig .
[ fig4 ] . again , these relations between the esr linewidth and the interplay of kondo effect , rkky interactions and anisotropy support the theoretical basis describing the spin dynamics in terms of a collective spin mode of kondo ions and conduction electrons.@xcite + next , we consider the linewidth behavior towards low temperatures . there ,
broadening contributions from magnetic fluctuations become important even far above the magnetic transition temperature because the resonance relaxation is very sensitive to the effect of critical spin fluctuations.@xcite the type of magnetic order underlying the magnetic phase diagram in fig .
[ pd ] depends on the co - content . as a consequence ,
characteristic differences in the low - temperature esr broadening can be identified for low and high co concentrations , roughly separated by the pronounced minimum of @xmath17 near @xmath89 . as can be seen in fig .
[ fig5 ] for almost all investigated co concentrations the esr line gets continuously narrowed down to the lowest accessible temperatures .
a considerable esr broadening toward low temperatures is only seen for @xmath92 .
however , the @xmath103 sample may also show a similar broadening at lower temperatures because the linewidth temperature slope @xmath98 is similar to the one for @xmath92 at @xmath130 k. the linewidth of the samples with co concentrations @xmath102 show deviations from the temperature linearity below @xmath90 k. + the sensitivity of the esr linewidth to the type of magnetic order in yb(rh@xmath107co@xmath108)@xmath3si@xmath3 strongly resembles the observations from a local gd@xmath4 esr in the heavy - fermion compound ce(cu@xmath1ni@xmath14)@xmath0ge@xmath0 .
there , the nature of magnetic order changes significantly from @xmath131 ( local moment type magnetism ) to @xmath132 ( itinerant heavy - fermion band magnetism).@xcite the temperature dependence of the gd@xmath4 linewidth could characterize the different ground states and , in particular , provide information about ce@xmath4 spin - fluctuations and the scattering of conduction electrons at the gd@xmath4 spins ( for which , moreover , non - fermi liquid behavior could be identified ) .
the effect of ce@xmath4 spin - fluctuations could be described by a @xmath133 dependence . in the case of yb(rh@xmath107co@xmath108)@xmath3si@xmath3
the yb@xmath4 spin fluctuations lead to an increase of the linewidth , as clearly visible for the data of @xmath92 . as shown by the solid line in fig.[fig5 ]
a @xmath133 dependence reasonably describes the data .
+ previous esr investigations of ybrh@xmath0si@xmath0 under pressure and with co - dopings @xmath13 have shown a close relationship between the @xmath6-factor and the magnetic susceptibility @xcite , pointing out a typical property of a _ local _ esr probe .
this feature was discussed in terms of a molecular field model for ybir@xmath0si@xmath0 in ref . and
for ybco@xmath0si@xmath0 in sec .
[ sec3a ] .
that means that internal magnetic fields lead to an effective @xmath6-factor @xmath134 which is shifted from the ionic @xmath6-factor @xmath135 depending on a molecular - field parameter @xmath136 : -factor and magnetic susceptibility @xmath137 with @xmath47-axis .
insets show the relation for @xmath48-axis .
dashed lines indicate a molecular - field relationship as discussed in the text . ]
.meanfield parameters of @xmath6 obtained from linear fits as shown by the dashed lines in fig . [ gchi ] . [
cols="^,^,^,^,^",options="header " , ] @xmath138\ ] ] for yb(rh@xmath107co@xmath108)@xmath3si@xmath3 the validity of this equation is nicely established as shown by the dashed lines in fig .
[ gchi ] with the values listed in table [ table1 ] .
the values @xmath139 successively decrease with increasing co - content .
discrepancies between these values and the values @xmath61 obtained from saturation magnetization could be due to van - vleck contributions to the susceptibility.@xcite the in - plane molecular - field parameter @xmath140 which is sensitive to the interactions for the @xmath141 components of the yb - moments shows negative values pointing to an antiferromagnetic exchange for the @xmath141 components . an _ abrupt _ change from @xmath142 for @xmath143 to @xmath144 for @xmath145 could be related to a corresponding change in the magnetic structure .
mssbauer results on pure ybco@xmath0si@xmath0 indicate the ordered moment to be within the basal plane.@xcite the out - of plane internal fields lead to a ferromagnetic @xmath6-shift as shown in the insets of fig .
[ gchi ] . there ,
positive and large molecular - field values @xmath146 describe the data .
the strongly enhanced value of @xmath146 for @xmath106 is a consequence of the small value of @xmath147 , since within a heisenberg picture @xmath136 can be related to the exchange parameter @xmath148 as @xmath149 .
hence , for @xmath99 @xmath150 is twice as large than for @xmath106 supporting an increase of the ferromagnetic exchange with increasing @xmath16 .
these results suggest that the exchange for the z and the x - y components of the local moment are predominantly ferromagnetic and antiferromagnetic , respectively .
the series yb(rh@xmath107co@xmath108)@xmath3si@xmath3 show well defined esr spectra with properties typical for local yb@xmath4 spins .
the esr @xmath6-factor shows a distinct and uniform variation with the co - content and no indications for a change in the symmetry of the ground state kramers doublet could be found . the linewidth evolution with the co - content clearly reflects variations of both the magnetic order and the kondo interaction . a strong reduction of the latter by co - doping leads to a line broadening .
this demonstrates that the kondo interaction crucially determines the esr linewidth in kondo lattices , especially in the presence of a strong magnetic anisotropy ( i.e. @xmath151 is large ) which is the case for yb(rh@xmath107co@xmath108)@xmath3si@xmath3 with a small co - content .
furthermore , our results suggest that ferromagnetic correlations may remain as the dominant mechanism for narrow esr lines when the kondo interaction is strongly suppressed and the magnetic anisotropy is sufficiently small , i.e. , at high co - contents of yb(rh@xmath107co@xmath108)@xmath3si@xmath3 .
this is a new aspect of the previously stated relevance of ferromagnetic correlations among 4@xmath8 spins for a narrow esr in kondo lattice systems .
+ the character of the linewidth temperature dependence towards low temperatures ( @xmath152 k ) is clearly different for low and high co - dopings .
this indicates changes in the magnetic fluctuations upon co - doping due to changes in the afm structure .
these changes could also be inferred from a change in the in - plane molecular - field around @xmath153 where the linearity between the magnetic susceptibility @xmath137 and the esr @xmath6-values changes slope . moreover
, this slope shows a different sign for @xmath154 and @xmath155 suggesting the exchange for the z and the x - y components of the moment to be predominantly ferromagnetic and antiferromagnetic , respectively .
we thank f. garcia , hans - albrecht krug von nidda and b.i .
kochelaev for fruitful discussions and acknowledge the volkswagen foundation ( i/84689 ) for financial support . | the evolution of spin dynamics from the quantum critical system ybrh@xmath0si@xmath0 to the stable trivalent yb system ybco@xmath0si@xmath0 was investigated by electron spin resonance ( esr ) spectroscopy . while the kondo temperature changes by one order of magnitude , all compositions of the single crystalline series yb(rh@xmath1co@xmath2)@xmath3si@xmath3 show well defined esr spectra with a clear yb@xmath4 character for temperatures below @xmath5k . with
increasing co - content the esr @xmath6-factor along the @xmath7-direction strongly increases indicating a continuous change of the ground state wave function and , thus , a continuous change of the crystal electric field .
the linewidth presents a complex dependence on the co - content and is discussed in terms of the co - doping dependence of the kondo interaction , the magnetic anisotropy and the influence of ferromagnetic correlations between the 4@xmath8 states .
the results provide evidence that , for low co - doping , the kondo interaction allows narrow esr spectra despite the presence of a large magnetic anisotropy , whereas at high co - concentrations , the linewidth is controlled by ferromagnetic correlations . a pronounced broadening due to critical correlations at low temperatures
is only observed at the highest co - content .
this might be related to the presence of incommensurate magnetic fluctuations .
[ published in phys .
rev .
b * 85 * , 035119 ( 2012 ) ] |
recent theoretical work suggests that feedback from agn plays an important role in establishing the present - day appearance of massive galaxies .
major galaxy mergers at high redshifts may trigger powerful starbursts , lead to the formation of elliptical galaxies , and account for the growth of supermassive black holes ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in this merger - driven evolutionary scenario ,
two gas - rich progenitors collide to form a completely obscured ulirg where substantial inflows of gas trigger starbursts and feed the central black hole(s ) . as the system evolves , the obscuring gas and dust begin to disperse , bringing about the formation of a dusty ir - excess quasar and eventually a `` naked '' optical quasar .
powerful winds , driven by either the starburst or the central black hole , seem to be required to quench star formation and to stop the growth of the black hole in order to explain the observed @xmath4 relation ( e.g. , * ? ? ?
there is growing support for the existence of this type of `` negative feedback '' in luminous late - stage mergers .
most galaxies with star formation rate ( sfr ) densities @xmath50.1 @xmath6 yr@xmath2 kpc@xmath7 show signs of outflow ( e.g. , * ? ? ?
* ; * ? ? ?
these starburst - driven winds are often seen in ulirgs , with mass outflow rates of @xmath30.15 times the sfr and velocities @xmath3100400 km s@xmath2 @xcite . in late - stage mergers and pure seyfert galaxies ,
however , these winds are sometimes driven by the agn , reaching outflow velocities @xmath51000 km s@xmath2 @xcite .
similar agn - driven outflows are also found in recently ( re)started radio galaxies ( e.g. , * ? ? ?
* ; * ? ? ?
several studies have also shown that these outflows are often multi - phased , with powerful molecular , neutral , and ionized outflows having all been detected in the same object ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in the present paper , we focus on the hi properties of the quest ulirgs and quasars , searching for the presence and the driving source ( agn or starburst ) of neutral outflows in merging systems as a function of merger stage .
this paper is organized as follows . in section [ sec : sample ] , we discuss how our hi sample is selected . in section [ sec : obs ] , we describe our observational setup , data reduction , and spectral analysis .
we then examine the hi profiles of our targets in section [ sec : profiles ] and the detection of polarized absorption ( including a more sophisticated spectral decomposition that boosts our number of absorption detections ) in section [ sec : polar ] .
finally , in section [ sec : origin ] , we discuss the possible origins of the hi outflows .
we conclude in section [ sec : sum ] with a summary of our findings . throughout this paper , we adopt @xmath8 = 71 km s@xmath2 mpc@xmath2 , @xmath9 = 0.27 , and @xmath10 = 0.73 @xcite .
in order to assess the physical conditions and the dynamical evolution of the neutral gas in late - stage merger systems , we selected targets from the quest sample of ulirgs and quasars that are bolometrically dominated by agn .
the details of the quest sample selection are outlined in previous quest papers ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
briefly , the original quest sample of ulirgs begins with the 1-jy sample , a flux - limited sample of 118 ulirgs selected from a redshift survey of the _ iras _ faint source catalog @xcite . to these ,
five well - studied local ulirgs from the revised bright galaxy survey ( rbgs ; * ? ? ?
* ) were added given the availability of archival data from multiple observatories ( see * ? ? ?
* ; * ? ? ?
* for details ) .
the original quasar sample contains 25 quasars , including 24 palomar - green ( pg ) quasars from the bright quasar sample @xcite and b2 2201 + 31a which satisfies the @xmath11 magnitude criterion of @xcite .
an additional nine pg quasars were added to the sample from archival data .
the selected quasars fall at the low end of the quasar @xmath11-band luminosity function .
the ulirgs and quasars in the quest sample are well - matched in redshift ( @xmath12 ) and in near - ir imaging show tidal features that indicate merger origins @xcite .
the entire quest sample spans the merger evolution sequence from binary starburst - dominated ulirgs , to completely coalesced agn - dominated ulirgs , to dust - enshrouded quasars , to optical quasars whose dust has been cleared .
our hi sample is a subset of the quest targets . to ensure quality spectra ,
we require @xmath13 ; our 27 targets evenly sample coalesced ulirgs ( 9 targets ) , far - infrared(fir)-strong quasars ( @xmath14 ; 10 targets ) , and fir - weak quasars ( 8 targets ) .
the two classes of quasars have similar spectral energy distributions representative of agn , with the fir - strong quasars having an additional contribution to their fir emission from intense star formation @xcite .
the morphologies of the quasar hosts are in agreement with the picture that fir - weak quasars are at a later evolutionary stage than fir - strong quasars , following the ( near ) cessation of star formation @xcite .
thus , our targets sample the later stages of merger evolution from coalescence to the aftermath of significant star formation .
table [ tab : sample ] summarizes the properties of our hi sample .
the targets were observed with the gbt spectrometer in the l band between 2011 august and 2012 march during the @xmath3160 hours of time allocated for gbt programs 11b075 and 12a075 ( pi : teng ) .
the data were collected in 10 minute on - off source pairs ( 5 minutes per position ) with 12.5 mhz bandwidth , two spectral windows centered at the same frequency ( 1420.4058 mhz ) , two linear polarization modes ( xx and yy ) , and nine sampling levels . to minimize the effects of radio frequency interference ( rfi ) on the data
, we used the notch filter for sources with heliocentric recessional velocities below 17000 km s@xmath2 .
the total on - source time for each galaxy was determined during the observing runs based on the strength of the spectral features relative to the noise or up to a maximum of three hours .
most sources were observed over multiple observing sessions .
the circumpolar starburst galaxy m82 was observed at the beginning of each observing session as a check for instrumental variations .
we also measured the continuum flux densities of our sample at 1.42 ghz using the autopeak procedure in order to check for time variability .
however , due to the plethora of continuum sources within the large gbt beam , the 5-@xmath15 confusion limit is @xmath3100 mjy at this frequency .
therefore , we were only able to obtain confident continuum measurements for five targets .
these values are in agreement with older values obtained with the very large array ( vla ) for their nuclear core flux densities @xcite , despite the much larger beam of the gbt . for the remainder of sample , we use published values of the core continuum flux densities as measured with the vla ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the spectra were reduced using gbtidl @xcite .
individual five - second records showing large broadband harmonic rfi were flagged , and persistent rfi spikes that appear in the data were replaced by interpolation across the affected channels for both spectral windows and polarizations . for each polarization ,
the data from the two spectral windows were averaged together , and a polynomial of order @xmath165 was fitted over a range of @xmath31000 km s@xmath2 on each side of a galaxy s systemic redshift in the hanning - smoothed and then decimated spectrum to subtract the baseline .
the same baseline - fitting regions and order of polynomial fit were used for both polarizations .
flux calibration was derived from simultaneous observations of an internal noise diode .
the baseline - subtracted and flux - calibrated scans for each polarization were accumulated and averaged for each galaxy . a fourth order boxcar smoothing
was then applied to the averaged data with decimation , resulting in a resolution of @xmath36 km s@xmath2 per channel .
the two polarizations were then averaged together to produce the full intensity spectra shown in figure [ fig : specs ] .
we ensured that our flux calibration is correct by comparing the full intensity profile of mrk 231 with that measured by the vla @xcite ; both profiles are identical
. discussions of individual objects in the appendix include comparisons with previous data , whose generally good agreement with our data in terms of velocity profile and flux validate our data reduction process .
table [ tab : sample ] also provides a list of some observational parameters for our galaxies .
nearly all of the galaxies were detected .
two galaxies , pg 0804 + 761 and f154620450 , had strong rfi spikes overlapping with the expected hi frequencies . at the limit of three hours on source , pg 1126041 and f212191757 were not detected , and f041032838 and f151301958 were marginally detected .
the detections of pg 1426 + 015 and pg 1617 + 175 are uncertain because their putative emission features have widths similar to those of broadness of the baseline ripples before baseline subtraction .
the individual hi profiles in the spectra of our sample galaxy are complex combinations of absorption and emission . for simplicity
, we have measured the emission and absorption features separately , but we note that this approach may slightly alter the final measured values of these quantities
. more complex decomposition of the emission and absorption is only possible for a few sources because of the degeneracy between emission and absorption , as discussed in section [ sec : decompose ] . for galaxies with significant emission ( peak - to - rms @xmath17 )
, we measure the central velocity ( @xmath18 ) , width ( @xmath19 , @xmath20 ) , and total velocity - integrated flux ( @xmath21 ) using the gbtidl task _ gmeasure_. in these instances , since there are no obvious companions enclosed within the gbt beam , the total flux is integrated between velocities at which the flux density reaches zero .
the central velocity is the heliocentric velocity measured at the midpoint of the 20% flux level .
the width of emission is measured at the 20% and 50% levels ( @xmath19 and @xmath20 , respectively ) .
the measurement errors in these quantities are derived using the expressions presented in @xcite following the methodologies of @xcite and @xcite .
the hi column density is calculated using expression 12.17 from @xcite for optically thin radiation .
assuming the gbt antenna temperature to flux density gain conversion of @xmath32 k jy@xmath2 : @xmath22 where @xmath21 is the velocity integrated hi flux in units of jy km s@xmath2 .
the neutral gas mass is calculated following @xcite : @xmath23 where @xmath24 is the recessional velocity corrected for the influence of the virgo cluster , the great attractor , and the shapley supercluster as calculated by the nasa extragalactic database ( ned ) .
all of these measured and derived quantities are listed in table [ tab : emission ] . only a few absorption features are significantly detected ( by the criterion that peak - to - rms @xmath25 ) . however , many of the absorption features are very broad but shallow , so the peak - to - rms ratio is not necessarily the best determinant of a detection .
we also consider a feature to be detected if its full width at half maximum ( fwhm ) spans at least 10 velocity channels and it has a peak - to - rms @xmath26 .
these absorption features are quantified using the gbtidl task _ fitgauss _ , with one gaussian component at a time added to a spectrum until the residuals are statistically featureless . with the exception of pg 1119 + 120 , where a narrow absorption feature is present within the emission profile ,
only the absorption features are modeled . for pg 1119 + 120
, the emission was first modeled with gaussians and subtracted from the data to define a baseline .
the absorption was then modeled from the emission - subtracted spectrum .
the gaussian models are plotted ( as components and as sums ) in figure [ fig : specs ] .
the gbtidl task measures the central velocity ( @xmath18 ) , fwhm , the `` peak '' flux density , and their measurement errors for each gaussian component .
the full velocity - integrated absorbed flux ( @xmath27 ) , optical depth ( @xmath28 ) , and column density ( @xmath29 ) are then calculated from the best - fit models .
we assume the hi absorbing screen covers the nuclear continuum source , in which case the optical depth profile is : @xmath30,\ ] ] where @xmath31 is the hi line profile as a function of velocity and @xmath32 is the continuum flux density @xcite .
the column density is then calculated as : @xmath33 where @xmath34 is the spin temperature . in this paper
, we will assume a spin temperature of 1000 k ( see section [ sec : mdot ] ) . for absorption features in pg 1119 + 120 ,
pg 1244 + 026 , and pg 1440 + 356 , the absorption troughs are deeper than their nuclear continua , suggesting that sources of continuum emission extend beyond the nucleus .
table [ tab : absorption ] summarizes the measured and derived values for all of the detected absorption features . as a reality check
, we inspected the data to confirm that the choice of polynomial order for the baseline fitting does not affect the detection of absorption or emission .
specifically , we examined whether polynomial orders of 3 or 5 can introduce ripples into the data that masquerade as absorption or emission features . after repeating the baseline subtraction for our data using a range of polynomial orders ( 15 )
, we confirm that the same spectral features are still detected , even for marginal detections such as that in f041032838 .
the measured widths and velocities are in agreement with those measured with the best - fit baseline to within the errors .
therefore , we are confident that even the weak detections noted as `` marginal '' in tables [ tab : emission ] and [ tab : absorption ] are real .
the individual hi profiles of our targets are discussed in the appendix . in general , we see hi trends along the merger sequence but no correlation with infrared luminosity .
all of the detected ulirgs show hi absorption , typically with multiple velocity components , many of which are narrow ( fwhm @xmath35 km s@xmath2 ) and some of which are redshifted .
only 2/7 ( 29% ) of the detected ulirgs have hi emission .
the different velocity absorption components have different origins and are not necessarily caused by outflows or infall ; kinematics of circumnuclear gas disks may mimic the properties of outflows or infall in single dish data .
in a couple of cases , such as ugc 05101 and mrk 273 , the absorption profiles are nearly symmetrical about zero velocity . for the type 2 agn ugc 05101
, the profile is likely from absorption due to a slightly inclined rotating hi disk superposed on an obscured nucleus .
this interpretation is consistent with co maps of ugc 05101 showing smooth , regular rotation of the molecular gas @xcite . on the other hand ,
the co velocity field of mrk 273 is very distorted @xcite , so it is unlikely that the gas is in well - ordered disk - like rotation .
the double absorption in hi seen in mrk 273 probably corresponds to the binary obscured nuclei in this object .
thus , co maps can help eliminate gas disks as the origin of the absorption features . however , only a few galaxies in our sample have readily available co maps so direct comparisons are difficult .
comparison of single dish co and hi profiles for the quest targets is on - going , but it is beyond the scope of the present paper . the more complex profiles , such as those seen in f15250 + 3608 , f051892524 , and f151301938 , may be interpreted as multiple infalling or outflowing components .
they may also result from the superposition of an emission profile and a broad absorption profile .
the latter case is discussed in more detail in section [ sec : decompose ] .
in contrast , the hi profiles of the fir - strong quasars are generally detected not only in absorption ( 88% or 7 of the 8 detected sources ) but also in emission ( 75% ) .
the absorption features are on average @xmath360 km s@xmath2 broader than those seen in ulirgs or approximately three times greater than the typical fwhm measurement error in the ulirgs . both redshifted and blueshifted absorption features are seen , although the systemic velocities of these systems are uncertain . optical emission lines may underestimate recessional velocity since some fraction of the line emission may arise from outflowing material . in the case of pg 0050 + 124
( i zw 1 ) , this effect causes a shift of @xmath3700 km s@xmath2 ( see appendix ) .
a similar situation could also hold for pg 1440 + 356 . in three cases ,
pg 1119 + 120 , pg 1244 + 026 , and pg 1440 + 356 , the minima of the absorption troughs exceed the nuclear continuum flux densities .
this discrepancy implies that off - nuclear emission , not accounted for in the nuclear fluxes of table [ tab : sample ] , also contributes to the continuum emission in these four sources .
given resolution limit of the vla core measurements of pg 1119 + 120 and pg 1440 + 356 , the off - nuclear emission must be at least @xmath32030 kpc from the core ; for pg 1244 + 026 , the separation from the core must be at least 3 kpc . while we can not formally rule out the possibility that these absorption features are absorption of hi against continuum from background quasars , this possibility is statistically unlikely .
another possibility is that the flux densities of the cores in these galaxies are time variable as exemplified by the variability of the core in pg 1351 + 640 ( see appendix ) .
the hi profiles of the fir - weak quasars are different from those of the ulirgs and fir - strong quasars .
hi is seen mostly in emission ( 88% or 7 of 8 detected sources ) , and only 25% ( 2/8 ) show broad absorption .
the emission profiles are very broad ( @xmath36500 km s@xmath2 ) and bumpy , perhaps beginning to resemble those of regularly rotating disks that typically display gaussian or double - horned profiles ; the measured widths and velocities are consistent with previous measurements of quasars ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
these properties will be revisited in section [ sec : decompose ] .
considering now our entire hi sample , the distribution of equivalent widths ( ew ; defined as the integrated flux density for each component divided by the continuum flux density of the core ) is plotted in figure [ fig : ew_dist ] . for this definition ,
the equivalent widths of the absorption components are negative .
it is clear that the strengths of the emission and absorption features progress along the merger sequence , with ulirgs having the weakest features , the fir - strong quasars having the strongest absorption features , and the fir - weak quasars having the strongest emission features . in figure
[ fig : fwhmvsfr ] , we compare the fwhms of emission and absorption with the star formation rates of their host galaxies . for emission ,
the fir - weak sources have the broadest widths , the ulirgs have the narrowest widths , and the fir - strong quasars bridge the gap between those two classes . in the ulirgs , there appears to be a weak correlation between the width of the absorption components and the sfr , reminiscent of a similar relationship seen in na i @xcite .
however , given that multiple absorption components are associated with a single system ( and thus a single sfr value ) , the validity of the connection between the sfr and the fwhm of the absorption components in ulirgs is uncertain . on the other hand ,
absorption fwhm is not at all related to sfr in the quasars , further suggesting that the agn play a role in these systems .
an unexpected outcome of our gbt program is that many of the absorption features show a dependence on polarization ( figure [ fig : polspecs ] ) . indeed ,
13 of our 23 gbt - detected targets show some profile difference in their two polarizations ( 1 ulirg , 7 ir - excess quasars , and 5 optical quasars ) . for gbt observations
, it is expected that there is a @xmath376% difference between polarizations in terms of normalization , but the observed spectral shapes should be the same in both polarizations ( r. maddalena , private comm . ) .
this expectation is confirmed by the fact that our m82 data remain in agreement after baseline subtraction .
polarization variation in our quest sample is not rfi - related , as we have removed time intervals where the data are affected by broadband rfi . since it is improbable that the hi screen itself is polarized ,
the underlying continuum emission must be polarized .
the most dramatic examples of polarized spectra are seen in pg 1211 + 143 and pg 1440 + 356 .
not only are these spectra polarized , but they are also time variable on timescales of @xmath330 minutes ( figures [ fig : pg1211var ] and [ fig : pg1440var ] ) .
in fact , of our radio - detected sample galaxies , time variability is only seen in sources that have significant absorption and spectra that are significantly polarized .
a possible concern is that since we observed these sources over several sessions , variability could be an instrumental effect .
however , comparing the different sessions data for m82 , we find that time variability and polarization are not seen beyond what is expected of typical gbt observations .
moreover , session - to - session instrumental variability would not explain the 30-minute variability seen within the same observing session .
time varying polarizations are not seen only in the targets with the brightest continua , so the variations are not likely to be due to instrumental systematics from poor cancellation of baseline ripples in the position - switching scans @xcite
. a different sort of instrumental effect may be leakage of linear polarization from the linearly polarized background continuum into the total intensity of the absorption profiles .
we can not yet exclude this scenario since these observations were not designed for polarimetry and our current data lack the cross - correlated products ( xy and yx ) that would be needed to confirm these time variations .
if these time variations are real , they could be caused by one of several processes .
first , they could be due to variations in the compact continuum sources behind the hi screen .
as the x - ray light curves of pg 1211 + 143 and pg 1440 + 356 show ( figure [ fig : xrayvar ] ) , the x - ray emitting central region can be highly variable on a timescale of about 30 minutes .
although not contemporaneous with our gbt data , these x - ray data suggest that the agn continuum can vary on a 30-minute time scale if the radio continuum is indeed related to the x - ray producing regions .
however , if the polarization - dependent variations are simply a result of the changing compact source continuum , then we would expect the continuum to vary in block and would observe a change in the equivalent width of the hi absorption , not the frequency - dependent variations we see in these targets .
second , variability could be a result of changes in the structure of the hi screen .
however , the timescale for the destruction of small hi clouds in the interstellar medium is on the order of a myr , depending on the exact properties of the interstellar medium causing the absorption ( e.g. , * ? ? ?
* ; * ? ? ?
* ) ; it is therefore unlikely that cloud destruction causes the observed short - term variations .
similarly , it is unlikely that the movement of hi clouds into and out of our line of sight to the continuum source can cause the very broad variations observed in these objects .
the time variability of the polarization - dependent hi absorption is easier to explain if the polarized continuum emission comes from both the nuclear source and an extended source ( e.g. , radio jets ) . in this picture
( figure [ fig : gbtlos ] ) , the hi screen is also extended , covering both the nuclear source and extended jets with kinematically distinct clouds .
as the central source fluctuates , the ratio of central - to - jet emission also changes , resulting in the observed wavelength - dependent polarization variations . in this scenario
, there is no need for the structure of the hi screen to change on a fast timescale .
regardless of whether the time variability is real , it is certainly the case that the spectra shown in figure [ fig : polspecs ] are polarized .
again , we checked for baseline - subtraction - related effects on the detection of polarized absorption and found that the choice of the polynomial order of the baseline fit matters little to the detection of these polarization - dependent absorption features .
the most obvious example of polarization not due to observational systematics is seen in the spectra of pg 0050 + 124 ( i zw 1 ) .
this hi profile is not polarized at relative velocities of 5001000 km s@xmath2 but becomes polarization - dependent between relative velocities 500 and + 500 km s@xmath2 ( figure [ fig : polspecs ] ) .
this behavior argues against attribution of apparent polarization dependencies to poor baseline subtraction or session - to - session variations in our gbt observations .
at least in this one case , the polarization dependence must be due to something physically happening at relative velocities between 500 and + 500 km s@xmath2 .
when we compare the hi profiles of each galaxy in the two separate polarizations , it appears that one polarization typically has higher flux than the other , with the exception of pg 1440 + 356 , in which there is a velocity shift instead .
the effect is not systematic in the sense that one polarization is always higher than the other ; each is nearly equally likely to be the polarization with the higher intensity . in some cases , particularly in pg 1351 + 640 and pg 1411 + 442
, the lower - intensity profile appears to be the same as the higher - intensity profile with the addition of a broad absorption component . to follow up on this intriguing hint , we produced the difference spectra ( the spectrum in the polarization with the more negative hi profile minus the spectrum in the polarization with the more positive hi profile ) shown in figure [ fig : diffspecs ] .
the absorption features in these spectra are then modeled the same way as before , using the gbtidl task _ fitgauss_. when these absorption models are included , many of the full intensity spectra that need multiple components to explain the blue- and red - shifted components can be explained simply by hi emission that is combined with a single broad absorption component .
this pattern is most obvious in pg 1351 + 640 and pg 1411 + 442 .
the results of this type of complex decomposition of the spectra are provided in tables [ tab : complex_emission ] and [ tab : complex_absorption ] .
similar decompositions can explain the shapes of the full intensity spectra of the ulirgs f051892524 and f151301958 , and the fir - strong quasar pg 1244 + 026 .
however , since those spectra are not polarized , we can not create difference spectra to model the broad absorption component .
task statistically prefers the multiple - component models to the single - absorption model for these spectra .
overall , with the exception of pg 1501 + 106 , the decompositions result in stronger features in both absorption and emission . given the results of this exercise , we find the instances of hi absorption ( emission ) to be 100% ( 29% ) in detected ulirgs , 100% ( 88% ) in fir - strong quasars , and 63% ( 100% ) in fir - weak quasars . the more complex decomposition of the spectra does not drastically change the distribution of ews ( figure [ fig : ew_dist_com ] ) compared to figure [ fig : ew_dist ] .
figure [ fig : fwhmvsfr_com ] provides a comparison of the fwhms of the emission and absorption features from the complex modeling with the star formation rates of these galaxies . as in figure
[ fig : fwhmvsfr ] , the fir - weak quasars have the broadest emission features and the ulirgs have the narrowest emission features . in terms of absorption
, there is no obvious relationship between the absorption feature widths and the star formation rates .
as before , the fir - weak quasars have the broadest absorption features , with the fir - strong quasars spanning the range of widths between fir - weak quasars and ulirgs .
as discussed in section [ sec : profiles ] , the absorption components are not necessarily caused by infall or outflow .
it is possible , as seen in ugc 05101 ( edge on ) and mrk 231 ( face on ) , that these features can be associated with the kinematics of gas disks .
a circumnuclear gas disk seen nearly edge on with a non - isotropic gas mass ( or background continuum ) distribution can mimic characteristics of infall and outflow .
unfortunately , our single dish data can not unambiguously exclude gas disks as the source of the features we observe , but it is unlikely that all of our targets with absorption features have gas disks that are seen precisely edge - on . keeping this caveat in mind ,
we assume in the following discussion an outflow origin only for blueshifted hi absorption features with hi - corrected heliocentric velocities ( @xmath38 for more details .
a remaining source of uncertainty in the interpretation of hi line profiles is the question of whether the midpoint of the 20% flux level is always a good approximation of the true systemic velocity .
emission from molecular lines is a more reliable tracer of systemic velocity than hi , so the ideal solution here would be acquisition of single - dish co spectroscopy for the full quest sample . unfortunately , these observations are still pending . ] ) below 50 km s@xmath2 @xcite .
a number of mechanisms have been proposed to explain such hi outflows .
they include adiabatically expanding broad emission line clouds ( e.g. , * ? ? ? * ) , starburst- and agn - driven galactic winds ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , and agn - driven jets interacting with the surrounding interstellar medium ( ism ; e.g. , * ? ? ?
these mechanisms predict different outflow locations ( from parsec to kiloparsec scales ) and geometries ( wide angled vs. collimated ) .
following @xcite and @xcite , the mass outflowing ( @xmath39 ) at a velocity @xmath40 is calculated using : @xmath41 assuming spherical geometry , mass conservation , and a wind that flows from a distance of @xmath42 to infinity .
the outflow is also assumed to subtend a solid angle of @xmath43 within which the gas has a covering fraction of 1 .
to facilitate comparisons with the values derived from young radio galaxies by @xcite , we use the same assumptions of @xmath44 and @xmath45 kpc . applying these assumptions and equation 5 to the blueshifted absorption components where @xmath46
, we find mass outflow rates @xmath39 ( wide ) up to more than 6000 @xmath6 yr@xmath2 ( summarized in table [ tab : outflows ] ) .
these values are much larger than those of @xmath3760 @xmath6 yr@xmath2 measured by @xcite and the range measured in ulirgs ( 13133 @xmath6 yr@xmath2 ; * ? ? ?
many of the mass outflow rates we infer seem unrealistically large ; such high rates have not been measured in other astrophysical objects .
these values do , however , depend on the assumptions we have made above .
the mass outflow rate depends linearly on the size scale of the outflow region and the derived hi column density .
recall from equation 4 that the column density depends on the spin temperature and optical depth profile , which in turn depends on the brightness of the background continuum .
there are thus three key factors that affect our estimates of a mass outflow rate : the spin temperature of the gas , the assumed distance , and the background continuum .
the assumption of 1000 k for the hi spin temperature seems to be reasonably conservative . in the vicinity of powerful agn ,
the extreme conditions of the gas and the presence of shocks can result in spin temperatures @xmath473000 k ( e.g. , * ? ? ?
* ) , although we note that absorption - line opacity will be dominated by the coldest material in any multiphase medium .
spatially resolved interferometric maps of several quasars suggest that the outflow regions have size scales @xmath30.21.6 kpc @xcite . adopting the smaller distance of 0.2 kpc instead of 1 kpc would reduce the derived mass outflow rates by a factor of 5 , more consistent with those derived for ulirgs .
finally , the estimated hi column density also depends strongly on the geometry of the source of continuum emission .
we have assumed that this emission comes entirely from the core . however , the absorption troughs in several quasars in our sample are deeper than the core fluxes .
this excess implies that off - nuclear emission also contributes to the observed continuum . in this case , @xmath28 in equation ( 3 ) is overestimated and so is @xmath29 , by a factor equal to the actual nuclear / total continuum flux ratio , although a larger source size may compensate to some extent .
if the off - nuclear continuum source is indeed physically related to the target and not a background quasar that happens to lie within the gbt beam , our polarized spectra give more credence to the idea that radio jets play an important role in driving hi gas outflows ( e.g. , * ? ? ?
several sources ( see appendix ) in our sample with polarized spectra have spatially resolved radio jets detected at higher frequencies .
the resolutions of these vla data imply that the jets are launched within @xmath30.40.9 kpc of the core . considering a jet - induced outflow scenario for the hi absorption features
, we again apply equation 5 to the blueshifted outflow components and assume @xmath48 kpc and a narrow opening angle of 5@xmath49 for the jet . under these new assumptions ,
the range of mass outflow rates is 0.231 @xmath6 yr@xmath2 ( see col .
( 8) of table [ tab : outflows ] ) .
we consider these values lower bounds on the actual mass outflow rates , while those derived using @xmath44 and @xmath45 kpc ( col .
( 7 ) of table [ tab : outflows ] ) are upper bounds . under the jet scenario
, the kinetic energy release rate can be calculated using the equation from @xcite : @xmath50 the kinetic energy in the outflow is on the order of @xmath51 to @xmath52 erg s@xmath2 for our sample .
the association of jets with the detected hi absorption features is also seen in the distribution of radio loudness of our sample .
only one quasar in our sample is considered radio - loud based on optical - to - radio flux ratios ( see appendix ) , but optical data may be subject to contamination from the host galaxy , resulting in an underestimate of radio loudness . using the 1.4 ghz to 210 kev ratio as an indicator of radio loudness and a radio - loud cutoff value empirically derived from 1600 agn @xcite , we find that many objects in our sample are radio - loud ( table [ tab : sample ] ) .
these results are consistent with those found in a recent study by @xcite that suggest higher detection rates of hi absorption toward extended compact steep - spectrum and gigahertz - peaked spectrum sources than toward the cores of radio galaxies . the environments in which the jets interact with the interstellar medium can be complex , and we may be seeing a strongly polarized continuum being absorbed selectively by multiple neutral hydrogen clouds , resulting in polarized hi absorption .
the fraction of radio - loud sources with associated hi absorption ( 58% ) is far higher than the @xmath310% observed in the general population of agn . given the high star formation rates of the ulirgs , the circumnuclear starburst can contaminate the 1.4 ghz measurements . using the 1.4 ghz to sfr correlation of @xcite
, we corrected the 1.4 ghz flux for star formation .
of the three ulirgs with blueshifted absorption and x - ray detections , all ( 100% ) are radio - loud .
similarly , of the six fir - strong quasars with detected hi outflows and x - ray emission , three ( 50% ) are considered radio - loud ; one of the two ( 50% ) fir - weak quasars with hi outflows and x - ray flux are radio - loud .
this pattern recalls an idea put forward by @xcite on the formation of radio galaxies to explain why radio - quiet agn outnumber radio - loud agn .
the authors suggested that major mergers , such as the progenitors of our sample of galaxies , result in coalesced supermassive black holes with high spin values due to the conservation of angular momentum .
substantial energy is then extracted from these fast - spinning black holes to drive jets , yielding radio - loud agn .
thus , younger radio galaxies may have more powerful jets and therefore stronger jet - induced outflows .
this scenario is consistent with the observations of @xcite , whose sample of young or recently restarted radio galaxies show similar hi outflows .
the @xcite picture is also consistent with theoretical models of @xcite that suggest the radio loud / quiet dichotomy in agn is due to the spin of the central black hole .
even though our sample with mass outflow rates is relatively small , we see some trends in our outflow measurements ( table [ tab : outflows ] and figure [ fig : fwzidist ] ) .
however , given the small number of detected hi outflows , particularly with only two measurements for the fir - weak quasars , we can not statically confirm that the distributions we see in figure [ fig : fwzidist ] is significant .
there is also no obvious relationships between mass outflow rate and the bolometric luminosity , agn luminosity , or radio loudness for our sample .
systems with deep absorption and outflows become more prevalent as we progress along the merger sequence from the newly coalesced ulirgs to the fir - strong quasars .
the trend between outflow kinematics and star formation rates among the ulirgs ( figures [ fig : fwhmvsfr ] and [ fig : fwhmvsfr_com ] ) suggests that the hi outflows in these objects are driven mainly by the starbursts , with a median mass outflow rate in the jet scenario ( @xmath39 ( jet ) @xmath31
@xmath6 yr@xmath2 ) similar to their sfrs and previous na i - based outflow measurements @xcite .
ulirgs also tend to have the narrowest absorption features . in the fir - strong quasar systems ,
the median mass outflow rate is @xmath39 ( jet ) @xmath35.8
@xmath6 yr@xmath2 , considerably larger than their median sfr of @xmath31.0 @xmath6 yr@xmath2 . while the median mass outflow rate of the fir - weak quasars is lower , @xmath39 ( jet ) @xmath30.3
@xmath6 yr@xmath2 , so is their median sfr ( @xmath30.6 @xmath6 yr@xmath2 ) .
thus , the mass - loading factor , _
i.e. _ , the ratio of mass outflow rate to sfr , is much greater in the fir - strong than the value of near unity derived in ulirgs .
the same value is much lower in the fir - weak quasars , but this is difficult to compare given that this is based on a single measurement .
however , the distinction between the ulirgs and fir - strong quasars further supports the idea that the agn is playing a role in driving hi outflows in the quasars before the end of large - scale star formation .
hi in emission is much more prevalent among the quasars than the ulirgs , and the hi emission profiles increasingly resemble those of regularly - rotating neutral gas disks as we move from fir - strong to fir - weak quasars .
the amount of hi seen in emission increases modestly from @xmath34@xmath5310@xmath54 @xmath6 among the fir - strong quasars to @xmath37@xmath5310@xmath54 @xmath6 among the fir - weak quasars .
these masses are consistent with those measured in low - redshift quasars @xcite and the high end of the gas mass range seen in nearby active galaxies @xcite .
we may be witnessing the creation of hi disks from residual neutral gas at the centers of merger remnants .
we have analyzed of gbt hi spectra of 27 ulirgs and quasars selected from the quest sample .
we find that : 1 .
as mergers evolve from fully coalesced ulirgs to fir - strong quasars to fir - weak quasars , the characteristics of their hi profiles also change from mostly absorption in the ulirgs to mostly emission in the fir - weak quasars , with the fir - bright quasars having intermediate properties .
the kinematics of these systems change from narrow absorption features likely due to outflows driven by starbursts in ulirgs , to broad absorption features from jet - driven outflows driven by agn in fir - strong quasars , and finally to the emission profiles more characteristic of gas in rotation in fir - weak quasars . 2 . a large fraction of the quasars exhibit polarization - dependent absorption features , with 78% of fir - strong quasars and 63% of fir - weak quasars showing these features compared with 13% in ulirgs .
some of these features vary on short time scales ( down to @xmath330 minutes ) .
neither time variability nor polarization - dependent spectral features are seen in contemporaneous observations of m82 ( as expected ) , reducing the likelihood that these effects are instrumental artifacts .
these features are likely due to hi absorption of polarized nuclear and off - nuclear continuum emission . the large fraction ( @xmath360% ) of radio - loud sources with hi outflow - associated absorption
suggests a close relationship between radio jets and the hi screen .
the polarization - dependent absorption features are most likely due to the emergence of radio jets in these systems , contributing to the continuum emission against which the absorption features are detected .
the @xmath360% radio - loud fraction of our sources with hi outflows is far higher than the @xmath310% seen in the general population of agn .
this distinction suggests that radio jets play an important role in shaping the environments in these systems , which may tract a transition phase in the evolution of gas - rich mergers into `` mature '' radio galaxies .
we are grateful to the anonymous referee for providing useful comments that improved the paper .
we thank the gbt operators and the green bank staff , especially ron maddalena , for support during this program ( 11b-075 , 12a-123 ) , as well as tim robishaw and lisa wei for useful discussions .
we also thank sheila kannappan and her team for allowing us to examine their proprietary data on pg 1501 + 106 .
the national radio astronomy observatory is a facility of the national science foundation operated under cooperative agreement by associated universities , inc .
we made use of the nasa / ipac extragalactic databased ( ned ) , which is operated by the jet propulsion laboratory , caltech , under contract with nasa .
we also acknowledge the usage of the hyperleda database ( http://leda.univ-lyon1.fr ) .
is supported by a nasa postdoctoral program ( npp ) fellowship . s.v .
acknowledges support from a senior npp award held at the nasa goddard space flight center and from the humboldt foundation which provided funds for a long - term visit at mpe in 2012 .
_ facilities : _ .
barvainis , r. , lonsdale , c. , & antonucci , r. 1996 , aj , 111 , 1431 bell , e.f .
2003 , apj , 586 , 794 blundell , k.m . & beasley , a.j .
1998 , mnras , 299 , 165 bothun , g.d . ,
heckman , t.m .
, schommer , r.a . , & balick , b. 1984 , aj , 89 , 1293 carilli , c.l . ,
wrobel , j.m .
, & ulvestad , j.s .
1998 , aj , 115 , 928 chandola , y. , gupta , n. , & saika , d.j .
2012 , mnras , in press ( arxiv : 1211.6852 ) condon , j.j . ,
hutchings , j.b . , &
gower , a.c .
1985 , aj , 90 , 1642 condon , j.j .
1998 , aj , 115 , 1693 condon , j.j . ,
cotton , w.d .
, & broderick , j.j .
aj , 2002 , 124 , 675 di matteo , t. , springel , v. , & hernquist , l. 2005 , nature , 433 , 604 downes , d. & solomon , p.m. 1998 , apj , 507 , 615 emonts , b.h.c .
2005 , mnras , 362 , 931 elvis , m. , marengo , m. & karovska , m. 2002 , apj , 567 , l107 feruglio et al .
a&a , 518 , l155 fischer et al .
a&a , 518 , l41 fouque , p. , durand , n. , bottinelli , l. , gouguenheim , l. , & paturel , g. 1990 , a&as , 86 , 473 genzel , r. et al .
1998 , apj , 498 , 579 haynes , m.p . & giovanelli , r. 1984 , aj , 89 , 758 heckman , t.m . , armus , l. & miley , g.k .
1990 , apjs , 74 , 833 heckman , t.m . ,
lehnert , m.d .
, strickland , d.k .
, & armus , l. 2000 , apjs , 129 , 493 heckman , t.m .
2002 , in asp conf .
254 , extragalactic gas at low redshift , ed .
j. s. mulchaey & j. stocke ( san francisco , ca : asp ) , 292 hinshaw , g. et al .
2009 , apjs , 180 , 225 ho , l.c . , darling , j. , & greene , j.e .
2008 , apjs , 177 , 103 ho , l.c . & kim , m. 2009 , apjs , 184 , 398 holt , j. et al .
2006 , mnras , 370 , 1633 hopkins , p.f .
et al . 2005 , apj , 630 , 705 hopkins , p.f .
2008 , apjs , 175 , 356 hutchings , j.b . ,
gower , a.c . , & price , r. 1987 , aj , 93 , 6 iwasawa , k. et al .
2011 , a&a , 428 , 137 kennicutt , r. c. 1998 , ara&a , 36 , 189 kim , d .- c .
& sanders , d. b. 1998 , apjs , 119 , 41 knig , s. , eckart , a. , garca - marn , m. , & huchtmeier , w. k. , 2009 , a&a , 507 , 757 krug , h.b .
2010 , apj , 708 , 1145 la franca , f. , melini , g. , & fiore , f. 2010 , apj , 718 , 368 leipski , c. , falcke , h. , bennert , n. , & huttemeister , s. 2006 , a&a , 161 , 172 marganian , p. , garwood , r.w . ,
braatz , j.a . ,
radziwill , n.m . , & maddalena , r.j . 2006 , astronomical data analysis software and systems xv ( asp conf .
351 ) , ed .
c. gabriel , c. arviset , d. ponz , & s. enrique ( san francisco , ca : asp ) , 512 martin , c.l .
2005 , apj , 621 , 227 martin , c.l .
2006 , apj , 647 , 222 mckee , c.f . & cowie , l.l .
1997 , apj , 215 , 213 morganti , r. , oosterloo , t. a. , emonts , b. h. c. , van der hulst , j. m. , & tadhunter , c. n. 2003 , apj , 593 , l69 morganti et al .
2004 , a&a , 424 , 119 morganti et al .
2005a , a&a , 439 , 521 morganti et al .
2005b , a&a , 444 , 9 morganti et al .
2007 , a&a , 476 , 735 morganti , r. 2011 , ( arxiv : 1112.5093 ) murray , n. , quataert , e. , & thompson , t.a .
2005 , apj , 618 , 569 nagashima , m. , inutsuka , s. , & koyama , h. 2006 , apj , 652 , l41 netzer et al .
2007 , apj , 666 , 806 oneil , k. 2002 , aspc , 278 , 293 oosterloo et al .
2000 , aj , 119 , 2085 piconcelli , e. et al .
2005 , a&a , 432 , 15 rafter , s.e . ,
crenshaw , d.m .
, & wiita , p.j .
2009 , aj , 137 , 42 rohlfs , k. , & wilson , t. l. 2000 , tools of radio astronomy / k. rohlfs , t.l .
new york : springer , 2000 .
( astronomy and astrophysics library , issn0941 - 7834 ) rupke , d. s. n. et al .
2002 , apj , 570 , 588 rupke , d. s. n. , veilleux , s. , & sanders , d. b. 2005a , apjs , 160 , 115 rupke , d. s. n. , veilleux , s. , & sanders , d. b. 2005b , apj , 632 , 751 rupke , d.s.n .
2005c , apj , 632 , 751 rupke , d.s.n . & veilleux , s. 2011 , apj , 729 , l27 sanders , d.b .
1998 , apj , 328 , l35 sanders d.b .
2003 , aj , 126 , 1607 schmidt , m. & green , r.f .
1983 , apj , 269 , 352 schneider , s.e . , helou , g. , salpeter , e.e . , & terzian , y. 1986 , aj , 92 , 742 schneider , s.e . , thuan , t.x . ,
magri , c. , & wadiak , j.e .
1990 , apjs , 72 , 245 schweitzer et al .
2006 , apj , 649 , 79 scoville , n.z .
2000 , aj , 119 , 991 springel , v. et al .
2005 , nature , 435 , 629 sturm et al .
2006 , apj , 652 , l13 sturm et al .
2011 , apj , 733 , l16 tchekhovskoy , a. , narayan , r. , & mckinny , j.c .
2010 , apj , 711 , 50 teng , s.h .
& veilleux , s. 2010 , apj , 725 , 1848 ulvestad , j.s . ,
wrobel , j.m . , & carilli , c.l . 1999 ,
apj , 516 , 127 ulvestad , j.s . , antonucci , r.r.j . ,
& barvainis , r. 2005 , apj , 621 , 123 veilleux , s. , kim , d.c . , & sanders , d.b .
2002 , apjs , 143 , 315 veilleux , s. , cecil , g. , & bland - hawthorn , j. 2005 , araa , 43 , 769 veilleux , s. et al .
2006 , apj , 643 , 707 veilleux , s. et al .
2009a , apjs , 182 , 628 veilleux , s. et al .
2009b , apj , 701 , 587 veilleux , s. 2012 , journal of physics : conference series , 372 , 12001 whiteoak , j.b .
& gardner , f.f .
1972 , a&a , 21 , 159 wei , l.h . ,
kannappan , s.j . ,
vogel , s.n .
, & baker , a.j .
2010 , apj , 708 , 841 wilson , a.s . , &
colbert , e.j.m .
1995 , apj , 438 , 62 wilson , c.d .
et al . 2008 , apjs , 178 , 189
in this appendix , we describe the hi spectral properties of our gbt sample .
when appropriate , we compare the present data with previous observations of these galaxies .
pg 0007 + 106 : : the only radio - loud object in our sample based on classic radio - to - optical criteria .
this source is classified as a triple system in the third zwicky catalog ( iii zw 2 ) , and the observed velocity range is consistent with the a ( @xmath55 km s@xmath2 ) and c ( @xmath56 km s@xmath2 ) components of the system .
the gbt detection is assumed to be that of the brightest component , a. the hi emission profile of the full intensity spectrum agrees with the old arecibo spectrum of @xcite .
the absorption feature , also noted by @xcite , is more prominent in the xx spectrum of the gbt data .
pg 0050 + 124 ( i zw 1 ) : : the heliocentric radial velocity listed in table [ tab : sample ] for this source is from optical spectroscopy by @xcite .
the optically - derived velocity is @xmath3700 km s@xmath2 lower than the hi - derived value of 18285 km s@xmath2 @xcite .
the optical emission lines may arise in part from outflowing gas , so the optically - derived value may is underestimate the actual systemic velocity of this object .
the shape of the full intensity profile ( figure [ fig : specs ] ) at 18285 km s@xmath2 agrees with those previously published by @xcite and @xcite , and matches expectations for a regularly rotating disk .
however , at lower velocities , the profile becomes polarized and a broad absorption feature is centered at the optically - derived velocity ( see figure [ fig : diffspecs ] ) .
f04103 - 2838 : : there is a marginal detection of an emission feature and a weak but broad blueshifted absorption feature present .
f05189 - 2524 : : no emission is seen in the data .
however , several absorption components are detected .
an alternative explanation for the spectral shape is the superposition of an emission profile with a broad absorption component .
pg 0804 + 761 : : strong rfi is present at the expected frequency of this galaxy .
it is therefore excluded from our analysis .
pg 0844 + 249 : : the broad emission profile is asymmetric .
although the profiles in the two different polarizations differ slightly , they have shapes similar to that observed by @xcite .
our integrated gbt flux density is larger than that measured by @xcite .
the emission feature redward of the peak may be due to a companion at @xmath57 km s@xmath2 .
ugc 05101 : : multiple absorption components are seen in the gbt data .
we interpret the hi absorption profile as a highly absorbed core coincident with an inclined regularly rotating disk .
pg 1119 + 120 : : this object was previously observed by @xcite and @xcite with arecibo .
the profiles of the older spectra resemble our gbt observation , although the narrow absorption feature is not apparent in the lower - resolution arecibo data .
there is no obvious detection of a companion in the gbt data near 16000 km s@xmath2 as suggested by @xcite .
high resolution vla maps at 4.8 ghz show a bent jet extending to the northeast @xcite . at the resolution of these vla maps ( @xmath30.75@xmath58 ) ,
the jet is resolved at a distance of @xmath30.72 kpc .
pg 1126 - 041 : : this target was statistically undetected after three hours on - source , although there is a very weak broad absorption feature centered near its systemic velocity .
pg 1211 + 143 : : this broad absorption line system was undetected by @xcite ; when it was observed by @xcite with arecibo , standing waves were present in the data due to the strong continuum .
pg 1229 + 204 : : this object was observed by @xcite .
the profile of the gbt spectrum roughly agrees with the earlier data .
pg 1244 + 026 : : a broad , blueshifted absorption feature is present . as for f051892524 , an alternative explanation for the spectral shape may be emission absorbed by a single broad absorption feature .
mrk 231 : : the deep absorption trough is associated with the face - on disk of the galaxy @xcite .
our gbt full intensity hi profile of this sources mostly agrees with these previous observations , but we detect an additional absorption feature centered at relative velocity @xmath59300 km s@xmath2 that was first suspected by @xcite .
this feature is slightly polarized , with the xx spectrum having lower intensity .
although the polarization is not as high as those seen in figure [ fig : polspecs ] , if we give the spectra of mrk 231 the same treatment detailed in section [ sec : decompose ] , we find that this feature has fwhm @xmath31000 km s@xmath2
corresponding to a mass outflow rate of @xmath30.3 @xmath6 yr@xmath2 in the jet scenario .
the polarized feature is thus likely associated with the radio jet @xcite .
the velocity of this jet - enhanced outflow is similar to the value measured using na i in the jet - disturbed region ( 1400 km s@xmath2 ; * ? ? ?
mrk 273 : : this source is the only binary in our sample ; its nuclear separation is @xmath3750 kpc @xcite .
x - ray observations by @xcite identified its southwestern nucleus as the seyfert 2 nucleus and suggested that the northern nucleus may host a heavily obscured agn .
the double - peaked hi absorption profile is very symmetric and is likely due to the two obscured nuclei .
there appears to be a weak absorption feature in the yy spectrum at 500 km s@xmath2 .
pg 1351 + 640 : : the xx spectrum appears to show emission , while yy shows broad absorption .
the hi absorption data may also be variable .
@xcite note a jet - like structure in their 4.8 ghz vla maps and time variability in the core flux density .
the resolution of the @xcite data ( @xmath30.56@xmath58 ) corresponds to a distance of @xmath30.91 kpc .
although on a different spatial scale , this structure is consistent ( similar position angle ) with a single - sided extension hinted at by the 8.4 ghz vlba snapshot data presented by @xcite .
this feature is not seen in deeper vlba imaging at multiple frequencies , but it should be noted that the data from @xcite are only in the left - circular polarization .
pg 1411 + 442 : : the spectrum of this object is very similar to that of pg 1351 + 640 .
emission is seen in the xx spectrum and a two - component absorption feature in the yy spectrum .
pg 1426 + 015 : : @xcite detect a fairly narrow ( @xmath60 ) and well - defined hi line centered at @xmath61 .
we have examined our spectrum at the corresponding frequency and see hints of emission over a wider range of velocities ; however , this feature has such low significance and is so sensitive to the details of baseline fitting we classify its detection status as `` uncertain . ''
pg 1440 + 356 : : the profile of the hi absorption feature depends slightly on the polarization .
exceptionally , the feature is redshifted by @xmath3300 km s@xmath2 with respect to systemic velocity rather than being blueshifted .
a possible explanation is that the systemic velocity taken from ned and based on optical emission lines underestimates the actual systemic velocity because the line - emitting gas arises in part from outflowing material .
another possibility is that we may be detecting infalling hi .
emission is also present in the blue half of the spectrum , although this is only seen at one of several epochs .
pg 1448 + 273 : : the weakly polarized spectra show blueshifted absorption .
pg 1501 + 106 : : a weak emission feature is present in the data .
a blueshifted absorption feature is also visible , particularly in the xx spectrum .
a lot of the data for this source were affected by small ripples in the baseline . as a reality check
, we compared our data with similar gbt data from program 10a-070 ( pi : kannappan ) .
the ripples also appear to be present in a large fraction of their data .
the persistent ripples could be due to harmonic waves from a strong rfi signal at a frequency that is slightly outside the detector frequency range .
due to the expected weakness of the features , we conservatively flagged all of the data that showed these ripples .
f15130 - 1958 : : multiple weak absorption components are present in this object .
f15250 + 3608 : : this is the only ulirg in our sample that shows distinct polarization - dependent hi absorption .
the spectra are very similar to those of pg 1351 + 640 and pg 1411 + 442 , where weak emission is seen in the xx spectrum and absorption in yy .
f15462 - 0450 : : strong rfi is present at the expected observed frequency of this galaxy .
it is therefore excluded from our analysis .
pg 1617 + 175 : : broad emission is detected in this object .
the profiles of the emission differ slightly in the two polarizations .
notably , the emission feature in the yy spectrum seems to be affected by an additional absorption feature ( figure [ fig : diffspecs ] ) .
f21219 - 1757 : : this object was not detected .
pg 2130 + 099 : : this target was observed by @xcite .
our full intensity profile of this source agrees with the older data from arecibo .
a blueshifted absorption feature is also seen in the xx spectrum .
@xcite detected what looks like jet - associated hot spots emanating from the core in 4.8 ghz a - array vla data .
the resolution of these vla data ( @xmath30.35@xmath58 ) implies that the jet can be resolved at a distance of @xmath30.42 kpc .
pg 2214 + 139 : : the hi spectrum of this object is dominated by a broad emission feature centered at systemic velocity . lccccccrccccccc pg 0007 + 106 & 00:10:31.0 & + 10:58:30 & 26783 & qso & weak & 12.23 & @xmath31.2 & 0.099@xmath62 & 2.4 & 2011 sept
2012 jan & 7894 & 6.11 & 0.77 & 3 + pg 0050 + 124 & 00:53:34.9 & + 12:41:36 & 17658 & qso & strong & 12.07 & 10.4 & 0.009@xmath62 & 5.0 & 2011 sept 2012 jan & 4815 & 5.78 & 0.94 & 3 + f041032838 & 04:12:19.4 & 28:30:25 & 35215 & ulirg & strong & 12.30 & 154.2 & 0.013@xmath62 & 3.4 & 2011 sept 2012 mar & 7452 & 6.43 & 0.88 & 3 + f051892524 & 05:21:01.4 & 25:21:45 & 12760 & ulirg & strong & 12.22 & 71.6 & 0.029@xmath62 & & 2011 sept
2012 jan & 12330 & 5.60 & 0.58 & 3 + pg 0804 + 761 & 08:10:58.6 & + 76:02:43 & 29979 & qso & strong & 12.08 & 0.3 & 0.004@xmath62 & 5.3 & 2011 aug & 860 & 6.23 & 2.97 & + pg 0844 + 349 & 08:47:42.4 & + 34:45:04 & 19187 & qso & weak & 11.44 & 0.8 & @xmath63 0.0025@xmath62 & @xmath633.3 & 2011 nov & 3153 & 5.83 & 0.96 & 3 + ugc 05101 & 09:35:51.6 & + 61:21:11 & 11802 & ulirg & strong & 12.05 & 110.2 & 0.170 & 3.0 & 2011 aug 2011 sept & 2151 & 5.57 & 1.46 & 3 + pg 1119 + 120 & 11:21:47.1 & + 11:44:18 & 15050 & qso & strong & 11.33 & 0.4 & 0.004@xmath62 & & 2011 oct 2012 mar & 3837 & 5.68 & 0.83 & 3 + pg 1126041 & 11:29:16.6 & 04:24:08 & 18575 & qso & strong & 11.52 & 2.5 & @xmath63 0.0025@xmath62 & @xmath634.6 & 2011 aug
2012 jan & 8871 & 5.81 & 0.54 & 3 + pg 1211 + 143 & 12:14:17.7 & + 14:03:13 & 24253 & qso & strong & 11.96 & 0.0 & 0.600 & 2.8 & 2011 aug 2012 mar & 5986 & 6.02 & 0.92 & 3 + pg 1229 + 204 & 12:32:03.6 & + 20:09:29 & 18890 & qso & weak & 11.56 & 0.4 & 0.003@xmath62 & 4.9 & 2012 mar & 3411 & 5.82 & 0.92 & 3 + pg 1244 + 026 & 12:46:35.2 & + 02:22:09 & 14443 & qso & strong & 11.02 & 1.5 & 0.002@xmath64 & 5.1 & 2011 aug 2012 mar & 3555 & 5.66 & 1.00 & 3 + mrk 231 & 12:56:14.2 & + 56:52:25 & 12642 & ulirg & strong & 12.60 & 174.2 & 0.309 & 2.7 & 2011 aug 2011 sept & 1257 & 5.60 & 1.52 & 3 + mrk 273 & 13:44:42.1 & + 55:53:13 & 11326 & ulirg & strong & 12.24 & 141.6 & 0.136 & 3.5 & 2011 aug 2011 sept & 1290 & 5.50 & 1.31 & 3 + pg 1351 + 640 & 13:53:15.8 & + 63:45:46 & 26442 & qso & strong & 12.04 & 26.1 & 0.034@xmath62 & 3.3 & 2011 sept
2012 feb & 8153 & 6.10 & 0.79 & 3 + pg 1411 + 442 & 14:13:48.3 & + 44:00:14 & 26861 & qso & weak & 11.78 & 0.0 & @xmath63 0.0025@xmath62 & @xmath634.6 & 2011 aug 2012 feb & 6346 & 6.12 & 0.92 & 3 + pg 1426 + 015 & 14:29:06.6 & + 01:17:06 & 25953 & qso & weak & 11.92 & 1.9 & 0.004@xmath62 & 5.0 & 2012 mar & 3153 & 6.08 & 1.09 & 5
+ pg 1440 + 356 & 14:42:07.4 & + 35:26:23 & 23700 & qso & strong & 11.80 & 10.3 & 0.005@xmath62 & 4.8 & 2011 aug 2011 dec & 6135 & 6.00 & 0.63 & 3 + pg 1448 + 273 & 14:51:08.7 & + 27:09:27 & 19487 & qso & weak & 11.43 & 0.1 & 0.003@xmath65 & 4.8 & 2011 aug 2011 feb & 8917 & 5.85 & 0.63 & 3 + pg 1501 + 106 & 15:04:01.2 & + 10:26:16 & 10919 & qso & strong & 11.33 & 0.0 & @xmath63 0.0025@xmath62 & @xmath635.7 & 2012 feb
2012 mar & 6540 & 5.53 & 0.77 & 3 + f151301958 & 15:15:55.2 & 20:09:17 & 32585 & ulirg & strong & 12.23 & 108.8 & 0.010@xmath62 & 2.4 & 2011 dec 2012 mar & 2666 & 6.33 & 1.25 & 3 + f15250 + 3608 & 15:26:59.4 & + 35:58:38 & 16535 & ulirg & strong & 12.12 & 109.6 & 0.018 & & 2011 aug
2011 nov & 2007 & 5.74 & 1.16 & 3 + f154620450 & 15:28:56.8 & 04:59:34 & 29917 & ulirg & strong & 12.28 & 168.7 & 0.013@xmath62 & 3.4 & 2011 dec & 860 & 6.23 & 4.14 & + pg 1617 + 175 & 16:20:11.3 & + 17:24:28 & 33708 & qso & weak & 11.74 & 4.0 & @xmath63 0.0025@xmath62 & & 2011 aug 2012 mar & 2547 & 6.38 & 1.59 & 3 + f212191757 & 21:24:41.6 & 17:44:46 & 33526 & ulirg & strong & 12.17 & 48.3 & 0.015@xmath62 & & 2011 oct
2012 mar & 8529 & 6.37 & 0.70 & 3 + pg 2130 + 099 & 21:32:27.8 & + 10:08:19 & 18880 & qso & strong & 11.77 & 0.2 & 0.007@xmath62 & 2.7 & 2012 mar & 2001 & 5.82 & 1.13 & 5 + pg 2214 + 139 & 22:17:12.2 & + 14:14:21 & 19715 & qso & weak & 11.77 & 0.0 & @xmath63 0.0025@xmath62 & @xmath635.0 & 2011 sept 2011 oct & 2121 & 5.85 & 1.32 & 3 + [ tab : sample ] lcccrccccrc pg 0007 + 106 & @xmath66 & @xmath67 & @xmath68 & 4.30 & 5.58 & @xmath69 & @xmath70 & @xmath71 + pg 0050 + 123 & @xmath72 & @xmath73 & @xmath74 & 7.00 & 7.45 & @xmath75 & @xmath76 & @xmath77 + f04103 - 2838 ( m ) & @xmath78 & @xmath79 & @xmath80 & 3.44 & 3.91 & @xmath81 & @xmath82 & @xmath83 + pg 0844 + 249 & @xmath84 & @xmath85 & @xmath86 & 8.83 & 9.20 & @xmath87 & @xmath88 & @xmath89 + pg 1119 + 120 & @xmath90 & @xmath91 & @xmath92 & 5.35 & 6.45 & @xmath93 & @xmath94 & @xmath95 + pg 1211 + 143 & @xmath96 & @xmath97 & @xmath98 & 3.85 & 4.18 & @xmath99 & @xmath100 & @xmath101 + pg 1229 + 204 & @xmath102 & @xmath103 & @xmath104 & 5.37 & 5.84 & @xmath105 & @xmath106 & @xmath107 + mrk 273 & @xmath108 & @xmath109 & @xmath110 & 5.94 & 4.53 & @xmath111 & @xmath112 & @xmath113 + pg 1351 + 640 & @xmath114 & @xmath115 & @xmath116 & 2.64 & 3.34 & @xmath117 & @xmath118 & @xmath119 + pg 1426 + 015 ( u ) & @xmath120 & @xmath121 & @xmath122 & 5.12 & 4.69 & @xmath123 & @xmath124 & @xmath125 + pg 1440 + 356 & @xmath126 & @xmath127 & @xmath128 & 3.01 & 4.81 & @xmath129 & @xmath130 & @xmath131 + pg 1448 + 273 & @xmath132 & @xmath133 & @xmath134 & 2.04 & 3.26 & @xmath135 & @xmath136 & @xmath137 + pg 1501 + 106 & @xmath138 & @xmath139 & @xmath140 & 2.58 & 3.37 & @xmath141 & @xmath142 & @xmath143 + pg 1617 + 175 ( u ) & @xmath144 & @xmath145 & @xmath146 & 6.94 & 4.36 & @xmath147 & @xmath148 & @xmath149 + pg 2130 + 099 & @xmath150 & @xmath151 & @xmath152 & 7.86 & 6.95 & @xmath153 & @xmath154 & @xmath155 + pg 2214 + 139 & @xmath156 & @xmath157 & @xmath158 & 11.22 & 8.53 & @xmath159 & @xmath160 & @xmath161 + [ tab : emission ] lcccrrcc + f051892524 & @xmath162 & @xmath163 & @xmath164 & 4.73 & 1.50 & 0.099 & 0.04 + ugc 05101 & @xmath165 & @xmath166 & @xmath167 & 4.59 & 3.24 & 0.040 & 0.54 + & @xmath168 & @xmath169 & @xmath170 & 6.70 & 1.09 & 0.059 & 0.18 + & @xmath171 & @xmath172 & @xmath173 & 3.59 & 1.32 & 0.031 & 0.22 + pg 1119 + 120 & @xmath174 & @xmath175 & @xmath176 & 5.06 & & & 0.02 + pg 1211 + 143 & @xmath177 & @xmath178 & @xmath179 & 40.95 & 5.73 & 0.065 & 3.07 + pg 1244 + 026 & @xmath180 & @xmath181 & @xmath182 & 3.05 & & & 0.13 + mrk 231 & @xmath183 & @xmath184 & @xmath185 & 11.07 & 2.15 & 0.056 & 0.64 + mrk 273 & @xmath186 & @xmath187 & @xmath188 & 8.71 & 5.60 & 0.088 & 0.74 + & @xmath189 & @xmath190 & @xmath191 & 8.11 & 2.66 & 0.081 & 0.35 + pg 1440 + 356 & @xmath192 & @xmath193 & @xmath194 & 15.27 & & & 0.57 + + pg 0050 + 124 & @xmath195 & @xmath196 & @xmath197 & 1.42 & 1.62 & 0.161 & 0.01 + f041032838 ( m ) & @xmath198 & @xmath199 & @xmath200 & 1.33 & 8.10 & 0.094 & 0.09 + f051892524 & @xmath201 & @xmath202 & @xmath203 & 2.22 & 1.03 & 0.046 & 0.03 + & @xmath204 & @xmath205 & @xmath206 & 1.88 & 1.20 & 0.038 & 0.03 + & @xmath207 & @xmath208 & @xmath209 & 1.81 & 0.90 & 0.037 & 0.03 + pg 1211 + 143 & @xmath210 & @xmath211 & @xmath212 & 1.24 & 0.08 & 0.002 & 0.04 + & @xmath213 & @xmath214 & @xmath215 & 2.68 & 0.09 & 0.004 & 0.05 + pg 1244 + 026 & @xmath216 & @xmath217 & @xmath218 & 1.52 & 59.07 & 1.407 & 0.08 + mrk 231 & @xmath219 & @xmath220 & @xmath221 & 2.06 & 0.19 & 0.010 & 0.06 + pg 1351 + 640 & @xmath222 & @xmath223 & @xmath224 & 1.94 & 1.02 & 0.046 & 0.03 + & @xmath225 & @xmath226 & @xmath227 & 1.23 & 2.88 & 0.029 & 0.09 + pg 1411 + 442 & @xmath228 & @xmath229 & @xmath230 & 2.81 & & & 0.14 + & @xmath231 & @xmath232 & @xmath233 & 1.84 & @xmath4739.78 & 1.139 & 0.06 + & @xmath234 & @xmath235 & @xmath236 & 1.90 & @xmath4728.08 & 1.217 & 0.04 + pg 1448 + 273 & @xmath237 & @xmath238 & @xmath239 & 2.70 & 55.58 & 0.829 & 0.12 + & @xmath240 & @xmath241 & @xmath242 & 1.18 & 10.27 & 0.283 & 0.03 + pg 1501 + 106 & @xmath243 & @xmath244 & @xmath245 & 2.77 & @xmath4783.12 & 1.884 & 0.12 + & @xmath246 & @xmath247 & @xmath248 & 1.28 & @xmath4717.46 & 0.498 & 0.04 + f151301958 ( m ) & @xmath249 & @xmath250 & @xmath251 & 1.66 & 3.86 & 0.232 & 0.03 + & @xmath252 & @xmath253 & @xmath254 & 1.40 & 6.10 & 0.192 & 0.05 + & @xmath255 & @xmath256 & @xmath257 & 1.40 & 6.56 & 0.191 & 0.05 + f15250 + 3608 & @xmath258 & @xmath259 & @xmath260 & 2.21 & 2.42 & 0.154 & 0.04 + & @xmath261 & @xmath262 & @xmath263 & 2.40 & 3.41 & 0.168 & 0.06 + [ tab : absorption ] lrrrrccccrc pg 0007 + 106 & @xmath264 & @xmath265 & @xmath266 & 5.74 & 7.41 & @xmath267 & @xmath268 & @xmath269 + pg 0050 + 123 & @xmath270 & @xmath271 & @xmath272 & 8.32 & 6.80 & @xmath273 & @xmath274 & @xmath275 + pg 0844 + 249 & @xmath276 & @xmath277 & @xmath278 & 10.65 & 10.06 & @xmath279 & @xmath280 & @xmath281 + pg 1119 + 120 & @xmath282 & @xmath283 & @xmath284 & 6.23 & 7.43 & @xmath285 & @xmath286 & @xmath287 + pg 1211 + 143 & @xmath288 & @xmath289 & @xmath290 & 13.10 & 13.97 & @xmath291 & @xmath292 & @xmath293 + pg 1351 + 640 & @xmath294 & @xmath295 & @xmath296 & 3.56 & 4.72 & @xmath297 & @xmath298 & @xmath299 + pg 1411 + 442 & @xmath300 & @xmath301 & @xmath302 & 3.22 & 3.19 & @xmath303 & @xmath304 & @xmath305 + pg 1448 + 273 & @xmath306 & @xmath307 & @xmath308 & 3.07 & 4.83 & @xmath309 & @xmath310 & @xmath311 + pg 1501 + 106 & @xmath312 & @xmath313 & @xmath314 & 3.22 & 4.53 & @xmath315 & @xmath316 & @xmath317 + f15250 + 3608 & @xmath318 & @xmath319 & @xmath320 & 5.44 & 3.67 & @xmath321 & @xmath322 & @xmath323 + pg 1617 + 175 ( u ) & @xmath324 & @xmath325 & @xmath326 & 9.23 & 5.81 & @xmath327 & @xmath328 & @xmath329 + pg 2130 + 099 & @xmath330 & @xmath331 & @xmath332 & 8.56 & 7.87 & @xmath333 & @xmath334 & @xmath335 + [ tab : complex_emission ] lccrrrcc pg 0007 + 106 & @xmath336 & @xmath337 & @xmath338 & 3.28 & 4.32 & 0.033 & 0.38 + pg 0050 + 124 & @xmath339 & @xmath340 & @xmath341 & 3.41 & 112.44 & 0.722 & 0.76 + pg 0844 + 249 & @xmath342 & @xmath343 & @xmath344 & 2.42 & & & 0.58 + pg 1119 + 120 & @xmath345 & @xmath346 & @xmath347 & 4.73 & & & 0.04 + & @xmath348 & @xmath349 & @xmath350 & 1.11 & 29.66 & 0.693 & 0.09 + pg 1211 + 143 & @xmath351 & @xmath352 & @xmath353 & 20.43 & 5.82 & 0.065 & 3.11 + & @xmath354 & @xmath355 & @xmath356 & 19.52 & 10.38 & 0.062 & 5.56 + pg 1351 + 640 & @xmath357 & @xmath358 & @xmath359 & 3.13 & 21.32 & 0.118 & 0.63 + pg 1411 + 442 & @xmath360 & @xmath361 & @xmath362 & 5.21 & & & 0.79 + pg 1448 + 273 & @xmath363 & @xmath364 & @xmath365 & 1.54 & 128.80 & 1.079 & 0.26 + pg 1501 + 106 & @xmath366 & @xmath367 & @xmath368 & 1.47 & @xmath369155.41 & 1.966 & 0.23 + f15250 + 3608 & @xmath370 & @xmath371 & @xmath372 & 2.73 & 60.56 & 0.386 & 0.92 + pg 1617 + 175 ( u ) & @xmath373 & @xmath374 & @xmath375 & 2.08 & & & 0.67 + pg 2130 + 099 & @xmath376 & @xmath377 & @xmath378 & 0.83 & 26.35 & 0.446 & 0.15 + & @xmath379 & @xmath380 & @xmath381 & 1.55 & 31.27 & 1.104 & 0.15 + [ tab : complex_absorption ] lrrrcccrrr + f04103 - 2838 ( m ) & 409 & @xmath382 & 81 & @xmath383 & @xmath384 & @xmath385 & 1089 & 4.4 & 41.6 + f05189 - 2524 & 374 & @xmath201 & 10 & @xmath386 & @xmath387 & @xmath388 & 97 & 0.4 & 40.3 + & 79 & @xmath204 & 12 & @xmath389 & @xmath390 & @xmath391 & 24 & 0.1 & 38.3 + 121@xmath392 & 0.5@xmath392 & + mrk 231 & 280 & @xmath219 & 2 & @xmath393 & @xmath394 & @xmath395 & 13 & 0.1 & 39.1 + f15130 - 1958 ( m ) & 847 & @xmath249 & 39 & @xmath396 & @xmath397 & @xmath398 & 818 & 3.3 & 41.9 + & 451 & @xmath252 & 61 & @xmath399 & @xmath400 & @xmath401 & 687 & 2.7 & 41.3 + 1505@xmath392 & 6.0@xmath392 & + f15250 + 3608 & 71 & @xmath402 & 606 & @xmath403 & @xmath404 & @xmath405 & & & + & 39 & @xmath389 & @xmath406 & @xmath407 & 605 & 1.6 & 40.8 + + pg 0050 + 124 & 32 & @xmath408 & 1124 & @xmath409 & @xmath410 & @xmath411 & 6072 & 24.3 & 41.6 + pg 1119 + 120 & 82 & @xmath412 & & @xmath413 & @xmath414 & @xmath415 & & & + & 17 & @xmath416 & 297 & @xmath417 & @xmath418 & @xmath419 & & & + pg 1211 + 143 & 270 & @xmath420 & 58 & @xmath421 & @xmath422 & @xmath423 & 392 & 1.6 & 40.6 + & 61 & @xmath424 & 104 & @xmath425 & @xmath426 & @xmath427 & 158 & 0.6 & 38.9 + 550@xmath392 & 2.2@xmath392 & + pg 1244 + 026 & 381 & @xmath428 & 152 & @xmath429 & @xmath430 & @xmath431 & 1450 & 5.8 & 41.4 + pg 1351 + 640 & 68 & @xmath432 & 213 & @xmath433 & @xmath434 & @xmath435 & & & + pg 1440 + 356 & 325 & @xmath192 & & @xmath436 & @xmath437 & @xmath438 & & & + pg 1501 + 106 & 140 & @xmath439 & @xmath47 1554 & @xmath440 & @xmath441 & @xmath442 & @xmath47 7813 & @xmath47 31.3 & @xmath47 41.6 + pg 2130 + 099 & 247 & @xmath443 & 313 & @xmath444 & @xmath445 & @xmath446 & 2453 & 9.8 & 41.5 + & 213 & @xmath447 & @xmath448 & @xmath449 & 1450 & 5.8 & 41.5 + + pg 0007 + 106 & 270 & @xmath450 & 43 & @xmath451 & @xmath452 & @xmath453 & 64 & 0.3 & 38.5 + pg 0844 + 249 & 158 & @xmath454 & & @xmath455 & @xmath456 & @xmath457 & & & + pg 1411 + 442 & 288 & @xmath458 & & @xmath459 & @xmath460 & @xmath461 & & & + pg 1448 + 273 & 106 & @xmath462 & 1288 & @xmath463 & @xmath464 & @xmath465 & & & + pg 1617 + 175 ( u ) & 95 & @xmath466 & & @xmath467 & @xmath468 & @xmath469 & & & + & 666 & @xmath455 & @xmath470 & @xmath471 & 64 & 0.3 & 38.5 + [ tab : outflows ] are labeled with m and u , respectively , after their names .
the x - axis is displayed as velocity relative to the heliocentric radial velocity of the galaxy given in table [ tab : sample ] and identified by the vertical dashed line .
the regions with hash marks are frequencies where the data are affected by persistent rfi spikes in some observing sessions , and the black regions are those where the data are affected by rfi spikes in all observing sessions .
the best gaussian fits to the absorption features are highlighted in blue , where the dotted lines are individual gaussian components ( when there are multiple overlapping components ) and the solid curves are the sums of these components .
the kinematics of these systems change from narrow absorption features likely due to outflows supported by the starbursts in ulirgs , to broad absorption features from jet - driven outflows supported by the agn in fir - strong quasars , and finally to the emission profiles more characteristic of gas in rotation in the fir - weak quasars .
, width=672 ] and [ tab : absorption ] .
the absorption components are shown as having negative ews .
targets with only upper limits to their continuum flux densities are excluded from this plot . along the merger sequence ,
the ulirgs have the weakest features in both emission and absorption , the fir - strong quasars have the strongest absorption features with moderate strength emission features , and the fir - weak quasars are dominated by the strongest emission features . , width=336 ] in table [ tab : emission ] ) and absorption ( bottom ; fwhm in table [ tab : absorption ] ) .
the colors represent ulirgs ( black ) , fir - strong quasars ( red ) , and fir - weak quasars ( blue ) .
the width of the hi emission in the top panel may be inversely correlated with the sfr .
the galaxies with the weakest star formation ( fir - weak quasars ) have the strongest emission features , while the galaxies with the strongest star formation ( ulirgs ) have the narrowest velocity widths .
the velocity widths of the fir - strong quasars fall between those of the ulirgs and the fir - weak quasars .
there also seems to be a weak correlation between the fwhm in absorption and the sfr for ulirgs , but this result is very uncertain as ulirgs tend to have multiple absorption components for a single sfr .
there is no apparent correlation between the sfr and the absorption parameters in the quasars , suggesting that the outflow - related absorption components are not the result of star formation .
, width=336 ] .
the colors represent the two different polarizations ( red : xx ; blue : yy ) .
polarized spectra are clearly seen in pg 0007 + 106 , pg 0050 + 124 , pg 1119 + 120 , pg 1211 + 143 , pg 1351 + 640 , pg 1411 + 442 , pg 1440 + 356 , and f15250 + 3608 .
weak polarized features may be present in pg 0844 + 349 , pg 1448 + 273 , pg 1501 + 106 , pg 1617 + 175 ( u ) , and pg 2130 + 099 .
the fir - strong quasars show the most distinct polarized features , perhaps reflecting the presence of polarized radio jets in these objects are contributing polarized continuum emission against which the absorption features are seen.,width=672 ] ) minus the spectrum in the polarization with the higher flux density ( pol@xmath472 ) , plotted in purple .
the black curve shows the gaussian model component(s ) best fitting to the difference spectrum as tabulated in table [ tab : complex_absorption ] .
the lower panel shows the residuals of the best fit model as applied to the pol@xmath473 spectrum and the pol@xmath472 spectrum to demonstrate the quality of the fit .
the colors in all of the bottom panels , the x - axis , and the hashed rectangles are the same as those in figure [ fig : polspecs ] .
the polarization dependence of the hi spectra can be explained by broad hi absorption affecting one polarization more than the other.,width=672 ] apart from the use of the complex fitting results when available ( tables [ tab : complex_emission ] and [ tab : complex_absorption ] ) .
the general trends with stage along the merger sequence are in agreement with those in figure [ fig : ew_dist].,width=336 ] apart from the use of the complex fitting results in section [ sec : decompose ] when available ( tables [ tab : complex_emission ] and [ tab : complex_absorption ] ) .
the general trends are in agreement with those in figure [ fig : fwhmvsfr].,width=336 ] | we present the results of a 21-cm hi survey of 27 local massive gas - rich late - stage mergers and merger remnants with the robert c. byrd green bank telescope ( gbt ) .
these remnants were selected from the quasar / ulirg evolution study ( quest ) sample of ultraluminous infrared galaxies ( ulirgs ; l@xmath0 l@xmath1 ) and quasars ; our targets are all bolometrically dominated by active galactic nuclei ( agn ) and sample the later phases of the proposed ulirg - to - quasar evolutionary sequence .
we find the prevalence of hi absorption ( emission ) to be 100% ( 29% ) in ulirgs with hi detections , 100% ( 88% ) in fir - strong quasars , and 63% ( 100% ) in fir - weak quasars .
the absorption features are associated with powerful neutral outflows that change from being mainly driven by star formation in ulirgs to being driven by the agn in the quasars .
these outflows have velocities that exceed 1500 km s@xmath2 in some cases .
unexpectedly , we find polarization - dependent hi absorption in 57% of our spectra ( 88% and 63% of the fir - strong and fir - weak quasars , respectively ) .
we attribute this result to absorption of polarized continuum emission from these sources by foreground hi clouds .
about 60% of the quasars displaying polarized spectra are radio - loud , far higher than the @xmath310% observed in the general agn population .
this discrepancy suggests that radio jets play an important role in shaping the environments in these galaxies .
these systems may represent a transition phase in the evolution of gas - rich mergers into `` mature '' radio galaxies . |
the burgeoning field of spintronics @xcite has created demands for entirely new classes of materials .
one such materials class combines both semiconducting and ferromagnetic properties .
magnetic semiconductors can help in the efficient injection of spin from a ferromagnetic contact to a semiconductor.@xcite the substitution of magnetic ions such as mn@xmath7 in traditional semiconducting materials such as gaas is a well - established route to making ferromagnetic semiconductors.@xcite however mn - substituted gaas as well as digital heterostructures such as gaas/0.5ml - mn @xcite have maximum curie temperatures of around 100 k or less . in the ideal case , we desire magnetic semiconductors with a magnetic transition above room temperature .
the ability to have both @xmath8- and @xmath9-type doping , a long spin - relaxation lifetime , and large carrier spin - polarization are other desirable features . proposed candidate materials
are strongly hole - doped wide band gap wurtzite semiconductors gan and zno , with magnetic transition metal _
_ t__m substituents on the cation sites.@xcite a number of recent studies on bulk and thin - film samples have ensued which suggest that zno:__t__m with _ _
t__m = co @xcite and with _ _
t__m = mn @xcite are ferromagnetic at room temperature , whilst others find spin - glass behavior .
@xcite results of our prior work on bulk samples of zno:__t__m systems suggest a complete absence of any magnetic ordering in well - characterized bulk samples.@xcite studies by kolesnik _
et al . _ and coworkers confirm these findings.@xcite careful density functional calculations by spaldin @xcite also point to difficulties in inducing ferromagnetism by transition metal substitution in zno . we find that while magnetic susceptibility indicates strong near - neighbor coupling in zno:__t__m ( _ _ t__m = mn and co ) , the mean - field coupling is exceedingly weak.@xcite it is the mean field term that would dictate magnetic ordering .
it is useful to temper expectation with the following observation : the few known ferromagnetic oxide semiconductors have rather low @xmath10s : bimno@xmath3 ( @xmath10 = 105 k ) , @xcite secuo@xmath3 ( @xmath10 = 26 k ) , @xcite ytio@xmath3 ( @xmath10 = 29 k ) , @xcite euo ( @xmath10 = 79 k).@xcite a promising alternative is to induce _ ferrimagnetism _ in a wide band gap semiconductor with _ two _ host cation sites : this would exploit the more natural tendency of spins in insulating oxides to anti - align and would require no conduction electrons . a natural structure for ferrimagnetism
is spinel , ab@xmath0o@xmath5 , where spins in the tetrahedral ( a ) and octahedral ( b ) cation sites are usually anti - aligned with respect to each other ; in addition , with two b sites for every a site , the possibility of a net ferrimagnetic moment exists .
figure [ fig : structure ] shows the cation network centered around a single tetrahedral a cation in the spinel structure .
the very high effective cation coordination in this structure type would mean that even at rather low concentrations of substituent transition metal ions , extensive magnetic coupling can be expected . in this contribution
, we have chosen spinel znga@xmath0o@xmath5 with a direct band gap of 4.1 ev as host.@xcite both znga@xmath0o@xmath5 as well as zno can be grown epitaxially on spinel mgal@xmath0o@xmath5 substrates @xcite , suggesting that the title compound can be incorporated in devices . by preparing nominal solid solutions [ znga@xmath0o@xmath1@xmath2[fe@xmath3o@xmath1@xmath4 ( @xmath11 ) ,
we have been able to induce magnetism , with strongly hysteretic behavior for the @xmath6 = 0.15 sample at 5 k. mssbauer spectra acquired at 293 k and 4.2 k on the @xmath6 = 0.15 sample suggests spontaneous magnetization at 4.2 k. the data also reveal the complete absence of fe@xmath12 in the sample , which supports our view of dilute ferrimagnetism rather than magnetism arising from a spurious source such as magnetite fe@xmath3o@xmath5 nanoparticles in a znga@xmath0o@xmath5 matrix .
diffuse reflectance uv / vis spectroscopy reveals that the direct band gap of znga@xmath0o@xmath5 is unaffected by substitution .
as in previous work on polycrystalline zn@xmath2m@xmath4o materials @xcite precursor oxalates zn@xmath2fe@xmath4(c@xmath0o@xmath5)@xmath132h@xmath0o were prepared with @xmath6 = 0.00 , 0.02 , 0.05 , 0.10 , and 0.15 by precipitation from aqueous solution .
polycrystalline samples with the nominal compositions [ znga@xmath0o@xmath1@xmath2[fe@xmath3o@xmath1@xmath4 ( @xmath14 were made by grinding together the corresponding oxalate with appropriate amounts of ga@xmath0o@xmath3 and decomposing in air at 1473 k for 18 hrs .
, with an intermediate regrinding step . for @xmath6 = 0.00 , the powder is white ; as @xmath6 increases , the powder colors range from beige to auburn .
beyond @xmath6 = 0.15 , samples require longer sintering times , suggesting a possible solubility limit in the vicinity .
step - scanned x - ray diffraction data on the powders were collected on a scintag x-2 diffractometer operated in the @xmath15-@xmath16 geometry .
transmission electron microscopy of the @xmath6 = 0.15 sample , in conjunction with energy dispersive x - ray spectroscopic ( edx ) analysis was carried out on a jeol jem 2010 microscope .
the sample powder was dispersed from a solvent on to a carbon - coated copper grid for tem studies .
uv / vis diffuse reflectance spectra were acquired on powders sprinkled on scotch tape .
magnetization data were collected on a quantum design mpms 5xl squid magnetometer operated between 2 k and 400 k. mssbauer spectra were collected at 293 k and 4.2 k using a constant - acceleration spectrometer equipped with a 1024 channel analyzer and operated in the time scale mode .
the @xmath17 source was 25 mci @xmath18co / rh .
the spectra were analyzed using the computer program effino.@xcite
x - ray diffraction data , plotted on a log scale in figure [ fig : x - ray](a ) , indicate a very small as - yet unidentified impurity at 34.2@xmath19 2@xmath15 in all the samples , including the non - magnetic host znga@xmath0o@xmath5 .
in addition , the @xmath6 = 0.15 sample has a small impurity of monoclinic ga@xmath0o@xmath3 ( pdf card 41 - 1103 ) .
other than these clearly non - magnetic impurities , the samples are clean showing only the spinel phase
. the diffraction profiles could be satisfactorily fit to the spinel crystal structure by the rietveld method using the xnd program.@xcite for all the samples , reliability factor ( @xmath20 ) was less than 7% .
the evolution of the @xmath21 cell parameter with @xmath6 is shown in figure [ fig : x - ray](b ) and is consistent with substitution of fe on the spinel znga@xmath0o@xmath5 lattice .
the line connects published cell parameters of znga@xmath0o@xmath5 @xcite and fe@xmath3o@xmath5.@xcite under the conditions of our preparation , fe@xmath3o@xmath5 is stable under the heat treatment conditions which we have employed,@xcite so we were initially confident in describing the fe substitution as a solid solution [ znga@xmath0o@xmath1@xmath2[fe@xmath3o@xmath1@xmath4 .
however , as we shall describe presently , mssbauer spectra indicate a complete absence of fe@xmath12 in the @xmath6 = 0.15 sample .
we do see a small broadening of the x - ray profiles with substitution characteristic of the increased number of elements in the crystal .
the x - ray form factors of fe , zn and ga are too similar to one - another for us to directly refine the relative amounts of these elements in the spinel crystal structure
. tem / edx analysis of the @xmath6 = 0.15 sample systematically indicated that the fe : zn : ga atomic ratio was close to 1:3:6 rather than the ratio calculated from the nominal composition which is 1:1.9:3.8 , or nearly 1:2:4 .
more importantly , the samples are homogeneous on the length scale of the edx spot size ( approximately 3 nm ) , consistently displaying the same ratios over several spots . across the substitution series
, the semiconducting nature of znga@xmath0o@xmath5 is retained as confirmed by uv / vis diffuse - reflectance spectroscopy .
diffuse reflectance uv / vis spectra across the series are shown in figure [ fig : optical ] .
znga@xmath0o@xmath5 is a direct band gap semiconductor,@xcite as confirmed by the sharp band - edge absorption . in our samples
we find the band edge is at 3.3 ev rather than the reported value of 4.1 ev .
however , it is well known@xcite that znga@xmath0o@xmath5 displays a very sensitive dependence of the optical absorption edge on composition , and a very small concentration of cation vacancies are sufficient to shift the edge to the red .
substitution by iron leaves the band - edge absorption unchanged , though new features associated with atomic transitions arise in the visible region .
the relative rise of these new features in the visible region of the spectrum results in an effective _ relative _ decrease in the intensity of the band - edge absorption .
figure [ fig : magnetism1](a ) shows the temperature dependence of the magnetization of three samples , @xmath6 = 0.05 , 0.10 , and 0.15 .
the magnetization has been scaled by the mole number of fe@xmath3o@xmath5 ; implying that in the absence of long - range interactions , traces for the three samples should collapse on a single curve .
we do not observe this even at 400 k. plots of @xmath22 as a function of @xmath23 are not linear for any of the samples below 400 k , also in accord with the existence of long - range interactions .
all three samples show a separation of the zfc and fc traces . this is clearly visible for @xmath6 = 0.10 and @xmath6 = 0.15 at about 100 k. at 5 k
, all three samples display hysteretic behavior as seen in figure [ fig : magnetism1](b ) .
none of the samples show magnetic saturation indicating that a certain fraction of the spins are `` free '' and are not involved in the ordering . if we assume that fe substitutes on znga@xmath0o@xmath5 as an alloy with fe@xmath3o@xmath5 , then the saturation magnetization should be 4 @xmath24 _ per _ fe@xmath3o@xmath5 .
if we take the minimum magnetic saturation value to be the point where the hysteresis loop closes , then for @xmath6 = 0.15 , this value is about 1 @xmath24 _ per _ fe@xmath3o@xmath5 at 5 k. therefore , to a good approximation , 25% of the substituted fe completely participate in the bulk ferrimagnetism .
figure [ fig : magnetism2 ] displays magnetization for @xmath6 = 0.15 as a function of the scaled field ( @xmath25 ) at three different temperatures , ( a ) 2 k , ( b ) 5 k , and ( c ) 200 k. at the two lower temperatures , hysteretic behavior is observed .
the hysteresis is lost by about 100 k , but even at 200 k , magnetization as a function of field indicates long range ordering .
the nature of the magnetization loops would suggest superparamagnetic behavior with a blocking temperature of around 100 k. however , superparamagnetism would imply that plots of @xmath26 as a function of @xmath25 would collapse on to a single -shaped trace above the blocking temperature .
we do not find this .
figure [ fig : magnetism2](d ) shows the temperature dependence of the magnetic coercivity of @xmath6 = 0.15 , which grows almost exponentially as the temperature is lowered . at 5
k , the coercivity is about 250 oe , to be compared with a value of 420 oe for fe@xmath3o@xmath5.@xcite mssbauer spectra were acquired at room temperature ( 293 k ) on @xmath6 = 0.05 sample and the @xmath6 = 0.15 sample . for both samples , the isomer shift with respect to fe metal was 0.32 mm s@xmath27 and the quadrupolar splitting was 0.52 .
neither sample showed , at 293 k , the 6-finger pattern characteristic of magnetic ordering . at 4.2 k , the @xmath6 = 0.15 sample displayed both a doublet corresponding to paramagnetic fe@xmath7 and a 6-finger pattern corresponding to magnetically ordered fe@xmath7 .
the internal field is between 250 and 510 oe , consistent with the magnetization data . from the mssbauer fit to the para and ferromagnetic phases at 4.2 k , a relative ratio of 10% ( paramagnetic ) to 90% ( ferromagnetic ) was obtained .
mssbauer data and the fit to different components , acquired on the @xmath6 = 0.15 sample are shown in figure [ fig : moess ] .
the relative distribution of fe@xmath7 on octahedral and tetrahedral sites could not be obtained from the present set of experiments .
experiments are planned in future with labeled @xmath18fe in order to obtain higher resolution data that will allow the fe ion site selection to be probed .
the title compounds have random substitution of ferric ions , presumably on both the tetrahedral and octahedral sites in the semiconducting host spinel structure .
since there are twice as many octahedral sites as there are tetrahedral sites , if the spins behave as they do in bulk fe@xmath3o@xmath5 , ferromagnetism can ensue .
we believe this to be at the origin of the observed magnetic behavior .
further mssbauer studies as well as powder neutron diffraction measurements are planned in order to better characterize the title compounds .
our results suggest that dilute _ _
ferri__magnetic semiconductors are worthy objects of study .
we thank gavin lawes for useful discussions and dan cohen for help with the optical measurements , and ombretta masala for help with tem studies .
a. s. r. is supported by the national science foundation igert program under the award dge-9987618 .
we gratefully acknowledge support from a care award ( sbb-0304a ) of the university of california / los alamos national laboratory . the work made use of facilities supported by the mrl program of the national science foundation under the award no .
dmr00 - 80034 . | solid solutions of nominal composition [ znga@xmath0o@xmath1@xmath2[fe@xmath3o@xmath1@xmath4 , of the semiconducting spinel znga@xmath0o@xmath5 with the ferrimagnetic spinel fe@xmath3o@xmath5 have been prepared with @xmath6 = 0.05 , 0.10 , and 0.15 .
all samples show evidence for long - range magnetic ordering with ferromagnetic hysteresis at low temperatures .
magnetization as a function of field for the @xmath6 = 0.15 sample is -shaped at temperatures as high as 200 k. mssbauer spectroscopy on the @xmath6 = 0.15 sample confirms the presence of fe@xmath7 , and spontaneous magnetization at 4.2 k. the magnetic behavior is obtained without greatly affecting the semiconducting properties of the host ; diffuse reflectance optical spectroscopy indicates that fe substitution up to @xmath6 = 0.15 does not affect the position of the band edge absorption .
these promising results motivate the possibility of dilute ferrimagnetic semiconductors which do not require carrier mediation of the magnetic moment . |
if the energy associated with the translational temperature of a gas is low compared to the relevant excitation energies and to the energy associated with the inverse observation time , then the gas can be considered as `` frozen '' and is analogous to an amorphous solid , or a glass .
such a frozen gas can be produced by laser cooling , and the atoms or molecules can be selectively placed in certain excited states .
when only a few states are involved , the system is analogous to a spin glass and can support polarization waves ( dipolar frenkel excitons ) that are analogous to spin waves . in this paper , we introduce a basic example of such waves and discuss their excitation . in laser experiments ,
the frozen gas is set up in a non - equilibrium electronic state , and the evolution of this state is measured after some elapsed time .
relaxation processes can be neglected for the duration of the experiment .
thus , the electronic state evolves towards dynamical equilibrium through the excitation of polarization waves in the random medium of the frozen gas .
the theory of this process must determine the spectrum of the polarization waves as well as their excitation probability during the transient evolution to dynamical equilibrium .
the two aspects of the problem will often be intermingled .
however , we have chosen to work on a system where they can be separated to some extent .
the system under consideration is fully described in an earlier paper by j. s. frasier , v. celli , and t. blum,@xcite henceforth to be referred to as fcb . in the present paper , we develop an improved theory of the particular case that is discussed in sec .
iii of fcb , under the heading `` sparse @xmath0 in a bath of @xmath1 . '' simply put , we consider one atom initially in state @xmath2 that is immersed in a frozen gas of atoms in state @xmath3 .
the @xmath2 state is connected to a @xmath4 state by a dipole matrix element @xmath5 , and an @xmath3 state is connected to a @xmath6 state by a dipole matrix element @xmath7 . the @xmath6 and @xmath4 states are singled out from all possible states , because the @xmath0 transition is resonant , i.e. , @xmath8 , and it is known that the transition rate peaks sharply at resonance.@xcite deviations from resonance are not considered in this paper .
the resonance condition is achieved by the application of a static electric field .
the @xmath0 process excites polarization waves described by @xmath9 transitions in the gas of @xmath3 atoms . typically , the number @xmath10 of @xmath3 atoms in the volume @xmath11 is such that @xmath12 @xmath13 .
the @xmath2 atoms are `` sparse '' if their number @xmath14 is such that @xmath15 , and in that case the result for @xmath14 atoms can be obtained by computing the result for one @xmath2 atom , averaging over distributions of the @xmath3 atoms , and then multiplying this result by @xmath14 .
we begin with the set of coupled differential equations in eq .
( 24 ) of fcb , fourier transformed in time : @xmath16 as in fcb , @xmath17 is the amplitude of the state in which the atom at the origin is in state @xmath2 and all other atoms are in state @xmath3 , while @xmath18 is the amplitude of the state in which the atom at the origin is in state @xmath4 , the atom at position @xmath19 is in state @xmath6 , and all other atoms are in state @xmath3 .
the quantity @xmath20 is the interaction potential @xmath21 evaluated at @xmath22 , and @xmath23 is the polar angle .
similarly , @xmath24 is @xmath25 evaluated at @xmath26 .
the angular factor present in @xmath27 and @xmath28 has been chosen for mathematical convenience : the actual angular factor is more complicated , as discussed in sec .
[ conc ] . in fcb ,
all the results depend only on an angular average of @xmath29 and , as is shown in the appendix of that paper , one can work with the effective angle - averaged potential @xmath30 . in the treatment presented here , however , the angular dependence plays a more important role . the potentials of eqs .
( [ v ] ) and ( [ u ] ) have particularly simple fourier transforms . defining @xmath31 we find , using the standard expansion of @xmath32 in spherical bessel functions and spherical harmonics,@xcite @xmath33 similarly , @xmath34 with these expressions we can obtain explicit solutions for the polarization waves in the continuum approximation , and we can then use these approximate solutions to describe the dynamics of a frozen gas state .
we change the sums in eqs .
( 1 ) to integrations over space , so that the evolution equations become @xmath35 where @xmath36 is now an averaged quantity . fourier transforming , we obtain simply @xmath37 the last equation shows that the frequencies of the polarization waves are @xmath38 independent of the magnitude of @xmath39 , but dependent on the direction , according to eq .
. this peculiar dispersion relation , which is analogous to well - known results in spin wave theory , is a consequence of the long range nature of the dipole - dipole interaction .
we expect it to be useful for values of @xmath40 up to the order of the inverse of the interatomic spacing.@xcite the solution of eqs .
( 8) gives @xmath41 where @xmath42 carrying out the angular integration , we find @xmath43m\left ( z\right ) , \label{meqn}\ ] ] where , leaving for sec .
[ capproxrandmed ] a discussion of the effective cutoff @xmath44 , @xmath45 and @xmath46 with @xmath47 a plot of @xmath48 and @xmath49 on the real axis is shown in fig .
[ figure1 ] . for large @xmath50 ,
@xmath51 as expected , @xmath52 is an analytic function in the upper @xmath50-plane , with a cut extending from @xmath53 to @xmath54 . on the real axis outside
the cut @xmath52 is real , and the spectral density on the cut is given by @xmath55 = \frac{\pi}{\sqrt{6}}\frac{z^{2}}{\sqrt{z+\frac{1}{2}}}.\ ] ]
to complete the evaluation of the self - energy @xmath56 of eq .
( [ meqn ] ) , it is necessary to introduce a cutoff in the @xmath40 integral .
we can do this in a way that makes the result applicable , at least approximately , to a random medium .
we know from sec .
ii of fcb that the solution of eqs .
( 1 ) for @xmath57 is of the form of eq .
( [ a0 m ] ) , with @xmath58 while in the continuum limit , using eq .
( [ vk]),@xcite @xmath59\frac{\left(\mu\mu^{\prime}\right)^{2}}{\omega}.\end{aligned}\ ] ] therefore , we see that if we make the substitution @xmath60^{2},\ ] ] then in the limit of @xmath28 going to zero our continuum result will agree with the result of fcb . applying this substitution , we find that eq .
( [ meqn ] ) becomes @xmath61^{2 } \nonumber \\ & = & -\frac{5m\left ( z\right)}{u_{0}}\sum_{l}v_{l}^{2}. \label{newmeqn}\end{aligned}\ ] ] we can now average over atom positions by the technique described in fcb , eq .
( 9 ) , with the result that , for @xmath62 , @xmath63 where , as in fcb eq .
( a2 ) , @xmath64 so that @xmath65 the integral in eq . ( [ a0av ] ) can be expressed in terms of the complementary error function.@xcite to develop approximations that automatically include a momentum cutoff we introduce the green s function of eq .
( 1b ) , which satisfies the equation @xmath66 then the solution of eqs .
( 1 ) is of the form of eq .
( [ a0 m ] ) , with the exact self - energy given by @xmath67 similarly , the on - site green s function , which clearly can not be treated in the continuum approximation , can be written as @xmath68}u_{pn } } , \label{gnn}\ ] ] while , for @xmath69 , @xmath70}u_{ln}g_{nn}+\sum_{m\neq l}g_{lm,[n]}u_{mn}g_{nn},\ ] ] where @xmath71}$ ] is the green s function for a medium where the @xmath72 atom is absent .
these equations are exact , but we now replace all occurrences of @xmath73 on the right hand side by @xmath74 , where @xmath75 denotes the average value of @xmath73 , which does not depend on @xmath72 . averaging and fourier
transforming the last equation and setting @xmath76\ ] ] we obtain @xmath77 similarly , writing the approximate version of eq .
( [ gnn ] ) in the form @xmath78}u_{pn } \right),\end{aligned}\ ] ] and expanding the second exponential , we obtain the self - consistent equation @xmath79 where @xmath80 is the fourier transform of @xmath81 this shows how a cutoff is automatically introduced into the theory . with these values of @xmath74 and @xmath82
we have approximately @xmath83 the self - consistency equation ( [ selfcon ] ) has been discussed extensively in the approximation where the contribution of @xmath84 is neglected ( i.e. , where contents of the last set of parentheses in eq .
( [ selfcon ] ) are replaced by unity).@xcite it is believed that this approximation is adequate when the polarization waves are localized in the random medium .
if @xmath74 is known as a function of @xmath85 , one can obtain @xmath86 directly .
comparing eq .
( [ gnn ] ) with @xmath87 we see that for an infinite medium @xmath88 it is also possible , in principle , to compute the average of @xmath89 for the frozen gas .
the result for @xmath90 with the @xmath56 of eq .
( [ newmeqn ] ) can be inverse fourier transformed and squared , and then the result must be averaged over all @xmath91 .
we have already discussed the cut of @xmath52 in sec .
[ contmodel ] , and one can see from eq .
( [ a0 m ] ) and the plots of fig . [ figure1 ]
that @xmath92 will have a pole at some @xmath93 , a pole at @xmath94 , and a pole at some @xmath95 .
therefore the inverse fourier transform can be computed by summing the residues of these poles and integrating around the cut , a process which is amenable to numerical techniques . however , the result is unreliable because the pole at @xmath94 is not really present in a random medium .
( in the continuum approximation , waves of zero frequency travel in the direction where @xmath96 and are not excited by the coupling @xmath27 , which vanishes at this angle . ) a similar calculation using the result of eq .
( [ m0best ] ) is very cumbersome . in practice ,
we have found it preferable to solve the equations ( 1 ) numerically for a given configuration of a few hundred atoms , and then average over an ensemble of configurations .
these results will be reported in a separate publication .
in the course of our theoretical investigation of the transient behavior of an optically excited gas , we have come across a particular class of polarization waves .
they are frenkel dipolar excitons , with the difference that the degeneracies of the @xmath3 and @xmath6 states have been lifted by the application of a static magnetic field .
first , we note that the spin - orbit coupling separates the @xmath97 states from the @xmath98 states .
then , considering the @xmath97 states , we see that the interaction between the @xmath99 and @xmath100 states is indeed proportional to @xmath101 , where @xmath102 is the angle between the interatomic vector and the applied magnetic field . in the actual experiments,@xcite
there is a spatially varying magnetic field in the magneto - optical trap , and the full analysis is much more complicated . in the simplest case
these waves are dispersionless and their frequency , given in eqs .
( [ uk ] ) and ( [ omegak ] ) , depends on the angle @xmath103 between the propagation vector and the field s direction according to @xmath104 . these waves have not ( to our knowledge ) been previously recognized , perhaps because their occurrence is confined to systems that can be described by the highly simplified eq .
( 1a ) , which additionally is treated in the continuum approximation .
we remark that the continuum approximation becomes exact in the long wavelength limit and is expected to have a wide range of validity because the dipole - dipole interaction is long - ranged .
the polarization waves also exist on a regular lattice , with a dispersion relation that can be calculated.@xcite in a frozen gas the excitation spectrum will spread out , but localization effects are probably small , again because of the long - range nature of the dipole - dipole interaction .
although in our model system almost all atoms are in an excited @xmath3 state , they could just as well be in the ground state .
what is important to the model is that this state couples effectively to a single @xmath6 state .
this is a good approximation for the rydberg atoms of the system we are modeling,@xcite but it is not a good approximation in general . also , the size of the dipole matrix element @xmath7 is particularly large for rydberg atoms . finally , we mention that the @xmath105 dipole - dipole interaction does not include retardation effects .
however , these are expected to influence the results only by inducing some dispersion at very long wavelengths . in this paper
we consider a particular mechanism of excitation of the polarization waves , namely , we allow the resonant transfer @xmath0 to create the initial @xmath6 state that then interacts with the @xmath3 states.@xcite however , other mechanisms of creation of the @xmath6 state would work just as well as long as a single @xmath6 state is created .
we have also assumed that if only a few @xmath6 atoms are created , then each can be treated independently in accordance with the sparse limit
. we outline ( but do not carry out in detail in this paper ) the calculation of the transient behavior involving the excitation of the newly - found polarization waves in a frozen gas .
while this allows us to establish contact with existing or currently planned experiments , one should perhaps think instead of more direct ways of observing the polarization waves .
acknowledgment is made to the thomas f. and kate miller jeffress memorial trust for the support of this research .
we also wish to thank p. g. wolynes for bringing to our attention related work done by him and his collaborators , and t. gallagher for many useful discussions .
j. d. jackson , _ classical electrodynamics _
( wiley , new york , 1975 ) , p. 767 . if @xmath106 cuts off at small @xmath107 , instead of being strictly proportional to @xmath105 , then the radial integral in eq .
( [ vk ] ) will give a @xmath40 dependence .
a spin - spin correlation function @xmath108 has the same effect .
see , for instance , d. e. logan and p. g. wolynes , `` self - consistent theory of localization in topologically disordered systems , '' phys .
b * 29 * , 6560 ( 1984 ) ; d. e. logan and p. g. wolynes , `` anderson localization in topologically disordered systems , '' phys . rev .
b * 31 * , 2437 ( 1985 ) ; k. s. schweizer , `` electronic absorption of frenkel excitons in topologically disordered systems , '' j. chem .
, 4638 ( 1986 ) . | laser experiments with optically excited frozen gases entail the excitation of polarization waves . in a continuum approximation
the waves are dispersionless , but their frequency depends on the direction of the propagation vector .
an outline is given of the theory of transient phenomena that involve the excitation of these waves by a resonant dipole - dipole transfer process . |
the theory of isothermic surfaces in conformal geometry has been the focus for intense research over the past years due to its relation with the theory of integrable systems ; see for instance @xcite and the literature therein . among isothermic surfaces darboux
@xcite distinguished the class of _ special isothermic surfaces_. these were introduced in connection with the problem of isometric deformation of quadrics and have been investigated by bianchi @xcite and calapso @xcite .
an isothermic immersion @xmath2 admits conformal curvature line coordinates @xmath3 for which its first and second fundamental forms read @xmath4 where @xmath5 is a nowhere vanishing smooth function and @xmath6 , @xmath7 denote the principal curvatures .
the immersion @xmath8 is called special of type @xmath9 if its mean curvature @xmath10 satisfies the equation @xmath11 where @xmath12 are real constants and @xmath13 being @xmath14 .
examples include surfaces of constant mean curvature ( cmc ) .
this article has its origins in the two seminal papers of bianchi @xcite on isothermic surfaces and deals with special isothermic surfaces of type @xmath0 , henceforth simply referred to as special isothermic .
these surfaces are invariant under the group of conformal transformations .
since the work of bianchi , it has been known that umbilic free special isothermic surfaces are conformally equivalent to cmc surfaces in 3-dimensional space forms .
examples of special isothermic surfaces with umbilic lines can be constructed by revolving elastic curves in the hyperbolic half - plane about the boundary at infinity ( cf .
section [ se.geom ] ) .
we prove that special isothermic immersions are in correspondence with the solutions @xmath15 of the partial differential equation defined by @xmath16 which only depends on the conformal class of the induced metric . ]
@xmath17 further , we show how the geometric properties of the darboux transformation of special isothermic surfaces apply to obtain analytic results for this equation . as for the study of darboux transforms of special isothermic surfaces ,
our work is related to that of hertrich - jeromin
pedit @xcite which discusses the case of cmc immersions in euclidean space and that of hertrich - jeromin musso nicolodi @xcite concerning immersions with cmc @xmath10 , @xmath18 , in hyperbolic 3-space . the special isothermic surface equation ( [ solitoneq ] )
can be reformulated as a zero - curvature equation @xmath19 = 0\ ] ] involving an auxiliary parameter @xmath20 .
this equation expresses the compatibility condition for a linear differential system @xmath21 in minkowski 5-space , where the lax pair matrices @xmath22 , @xmath23 take values in @xmath24 ( cf .
section [ se.lax ] ) . in this representation , for every solution @xmath25 of ( [ solitoneq ] ) , the spectral parameter @xmath20 describes a family of special isothermic surfaces which are second order conformal deformations of each other and are obtained by integrating the loop of @xmath24-valued 1-forms @xmath26 .
the linear system ( [ linearsys ] ) have the conserved quantities @xmath27 the classical darboux transforms of a special isothermic immersion with potential @xmath25 are constructed from the solutions @xmath28 of the linear system ( [ linearsys ] ) satisfying @xmath29 ( cf . @xcite and the recent papers @xcite ) . moreover ,
if @xmath28 satisfies the additional constraint @xmath30 , the darboux transforms of special isothermic surfaces are still special isothermic .
analytically , this is equivalent to the statement that if @xmath25 is a solution of ( [ solitoneq ] ) and if @xmath28 is a solution of the linear system satisfying @xmath31 , then @xmath32 is again a solution which can be regarded as a bcklund transformation of ( [ solitoneq ] ) .
this description furnishes a procedure for generating new solutions of ( [ solitoneq ] ) by solving ( [ linearsys ] ) . in section [ se.superpo ]
we prove a superposition formula for the solutions of ( [ solitoneq ] ) .
namely , if @xmath25 is a known solution and @xmath33 , @xmath34 are solutions respectively generated from @xmath25 by solutions @xmath28 and @xmath35 of ( [ linearsys ] ) corresponding to different values of the spectral parameter , then @xmath36\ ] ] represents a new solution of ( [ solitoneq ] ) .
this formula , which geometrically amounts to the bianchi permutability theorem for special isothermic immersions , shows that , after the first step , the procedure for obtaining new solutions can be carried out without the quadratures associated with the bcklund transformation ; it also provides a method for obtaining the multisoliton solutions of equation ( [ solitoneq ] ) by algebraic means only . as an application
, we compute the 1 and 2-soliton solutions arising from the trivial solution @xmath37 of ( [ solitoneq ] ) with @xmath38 .
the bcklund transformation for the fourth - order equation defining isothermic immersions the so - called calapso equation as well as its 1-solitons have been previously considered in @xcite and have been used by bernstein @xcite to obtain explicit examples of non - special isothermic tori with spherical curvature lines ( cf . examples in section [ se.baeck ] )
let us begin by recalling some basic facts .
the one - point conformal compactification @xmath39 of euclidean space is classically realized as the projectivization of the light cone @xmath40 of minkowski 5-space @xmath41 with lorentz scalar product @xmath42 : @xmath43\in \p^4 : \langle x , x\rangle=0 \right\}.\ ] ] we consider linear coordinates @xmath44 such that @xmath45 and identify @xmath46 with @xmath47 $ ] by means of the conformal map @xmath48\in \p [ \mathcal{l } ] , \quad j(\infty)=\left[\left(0,0,0,0,1\right)\right].\ ] ] the linear action of the pseudo - orthogonal group @xmath49 of ( [ innerproduct ] ) descends to a transitive action on @xmath46 by conformal ( mbius ) diffeomorphisms . in this model for @xmath46 , the de sitter space @xmath50 parametrizes the 2-spheres in @xmath46 by @xmath51 and @xmath52 acts transitively on the set of 2-spheres .
a _ mbius frame _ is a basis @xmath53 of @xmath54 such that the vectors form the columns of a matrix of @xmath52 .
geometrically , the unit space - like vectors @xmath55 , @xmath56 , @xmath57 represent 2-spheres which intersect orthogonally , and @xmath58,[b_4]\in \mathcal{m}$ ] their intersection points .
regarding @xmath59 as @xmath60-valued functions defined on @xmath52 , there are unique 1-forms @xmath61 such that @xmath62 and satisfying the structure equations @xmath63 @xmath64 is the maurer - cartan form of @xmath52 with values in the lie algebra @xmath65 of @xmath52 .
the set @xmath66 of all @xmath67 such that @xmath68 can be given a lie group structure . for
, let @xmath69 and define @xmath70 the mapping @xmath71 is a smooth embedding which induces a lie group structure on @xmath66 .
the group operation is given by @xmath72 in particular , @xmath73 and @xmath74 .
throughout the paper a solution @xmath25 of equation ( [ solitoneq ] ) will be referred to as a _ wave potential _ with _ character _ @xmath75 .
let @xmath76 be a simply connected domain with coordinates @xmath77 , and let @xmath78 be a nowhere vanishing smooth function . for @xmath79 , we define @xmath80 by @xmath81 a straightforward calculation gives : equation ( [ solitoneq ] ) is equivalent to the matrix lax equation @xmath19 = 0 , \ ] ] where @xmath82 are as defined above .
this is the compatibility condition for the linear system @xmath83 in minkowski 5-space .
the linear system ( [ darbouxsys ] ) is referred to as the @xmath84-_system _ ) is gauge equivalent to a special case of the system considered by darboux and bianchi for the construction the darboux transformation @xcite ( see also section [ se.baeck ] ) ] associated to @xmath25 .
the @xmath65-valued one - form @xmath85 satisfies the maurer - cartan equation @xmath86 and can consequently be integrated to a frame @xmath87 , uniquely determined up to left multiplication by a constant element of @xmath52 , such that@xmath88 we call @xmath89 a _ normal frame field _ for the wave potential @xmath25 with _ spectral parameter _ @xmath20 . because of ( [ structureequation ] ) , the components @xmath90 of any constant vector @xmath91 with respect to a normal frame @xmath89 provide a solution @xmath28 of the @xmath92-system . from ( [ structureequation ] )
it also follows that @xmath93 is a constant vector ; we call @xmath94 the _ pointing vector _ of the normal frame @xmath89
. the _ kinetic energy _ @xmath95 and the _ linear momentum _
@xmath96 of a solution @xmath28 of the linear system ( [ darbouxsys ] ) are defined by @xmath97 if @xmath98 is the initial condition of @xmath28 with respect to the normal frame @xmath89 , then @xmath99 and @xmath100 .
thus , @xmath101 and @xmath102 are two first integrals of the @xmath92-system .
let @xmath103 be an isothermic immersion with conformal principal coordinates @xmath3 and let @xmath104 be its fundamental forms , where @xmath5 is a nowhere vanishing smooth function and @xmath6 and @xmath7 are the principal curvatures .
recall that the notion of an isothermic immersion is conformally invariant , that is , if @xmath8 is an isothermic immersion and @xmath105 is a conformal diffeomorphism , then @xmath106 is also isothermic @xcite . in the following
we will not make any distinction between isothermic immersions in @xmath107 or in @xmath46 .
following bianchi @xcite we set @xmath108 and give the following @xmath8 is called _ special isothermic _ of type @xmath9 if there exist real constants @xmath12 such that @xmath109 we shall assume that @xmath110 is _ generic _ ,
i.e. , either one of the following conditions hold : @xmath111 , or @xmath112 and @xmath113 , or @xmath114 and @xmath115 , or else @xmath116 . next , let @xmath117 be the _ calapso potential _ of the isothermic immersion @xmath8 .
we are now in a position to state : let @xmath110 be a special isothermic immersion of type @xmath0 . then the calapso potential @xmath25 is a wave potential with character @xmath118 .
the gauss and codazzi equations @xmath119 @xmath120 imply @xmath121 and @xmath122 then , ( [ specialeq ] ) becomes are related to the christoffel transformation of @xmath8 .
thus , from ( [ varspecialeq ] ) it follows that the christoffel transform of an isothermic immersion of type @xmath0 is special of type @xmath123 , see also @xcite . ]
@xmath124 since @xmath125 equation ( [ varspecialeq ] ) also reads @xmath126 the derivative of ( [ varspecialeq ] ) with respect to @xmath127 and @xmath128 yields @xmath129 when @xmath130 the last two equations imply @xmath131 this equation combined with ( [ varhlm ] ) and ( [ gradtheta ] ) gives the result .
a similar argument applies in the other cases . in the setting of the previous proposition ,
notice that when @xmath132 the immersion @xmath8 has constant mean curvature .
when instead @xmath133 and @xmath134 the calapso potential @xmath135 is a function of the variable @xmath128 alone and @xmath8 is an isothermic canal surface . by a classical result of darboux @xcite ,
@xmath8 is then conformally equivalent to either a cone , a cylinder , or a surface of revolution .
if @xmath8 is a surface of revolution , one can prove that the profile curve lies in the hyperbolic half - plane , that its arclength is proportional to @xmath128 , and that its curvature is parametrized by the calapso potential .
further , equation ( [ solitoneq ] ) reduces to the euler - lagrange equation of the total square curvature functional @xmath136 on smooth curves @xmath137 with fixed length .
therefore , the surface defined by @xmath8 is obtained by revolving an elastic curve ( possibly a free elastic one ) in the hyperbolic half - plane about the boundary at infinity .
this description indicates , in particular , how to construct special isothermic surfaces with umbilic lines ( see also babich bobenko @xcite ) recall that rotational willmore isothermic surfaces arise from free elastic curves in the hyperbolic 2-plane @xcite .
see also langer
singer @xcite and pinkall @xcite . for the case of willmore canal surfaces
we refer to @xcite . as for the converse ,
consider first the following : for any @xmath138 , let @xmath139 be the conformal map defined by @xmath140 , where @xmath141 $ ] denote the canonical projection .
let @xmath87 be a normal frame field with wave potential @xmath25 and character @xmath75 , and let @xmath138 be a constant vector .
then , @xmath142 is a special isothermic immersion of type @xmath143 with calapso potential @xmath25 .
let @xmath28 be the solution of the @xmath92-system with initial condition @xmath144 , that is @xmath145 .
then , the fundamental forms of @xmath146 can be read off the maurer
cartan form of the euclidean frame @xmath147 .
these are computed to be @xmath148 this implies that @xmath149 is an isothermic immersion and that @xmath150 next , by using the constraint @xmath151 and the conservation of the linear momentum @xmath100 , a direct computation shows that @xmath152 thus @xmath8 is special isothermic of type @xmath143 .
the solution @xmath28 in the proof of the theorem can be expressed in terms of the euclidean invariants of @xmath8 by the following formulae : @xmath153 given an umbilic free immersion @xmath110 , there exist a canonical frame field @xmath154 along @xmath8 : the _ central frame _ field of the immersion .
the construction of such a frame is due to bryant @xcite . if @xmath8 is special isothermic of type @xmath0 with calapso potential @xmath25 , then @xmath155 is exactly the normal frame field with wave potential @xmath25 and spectral parameter @xmath156 .
let @xmath158 be a wave potential with character @xmath75 .
a solution @xmath28 of the linear system ( [ darbouxsys ] ) satisfying @xmath31 is said to be an _
@xmath20-system of transforming functions_-transforming functions do exist only if @xmath159 . if @xmath160 , then the set of @xmath20-transforming functions with potential @xmath25 is a @xmath161-dimensional cone .
if @xmath162 , the transforming functions are of the form @xmath163 for a constant @xmath164 .
observe that the last component @xmath165 of a system of transforming functions never vanishes . if , in addition , the character @xmath75 is positive , then also the third component @xmath166 never vanishes . ] for the potential @xmath25 .
the corresponding _ bcklund transform _ is defined by @xmath167 the bcklund transform @xmath168 is a wave potential with character @xmath75 .
let @xmath169 .
the first derivatives of @xmath170 are given by @xmath171,\quad \tilde{\phi}_{y}=\frac{v^3}{v^4}\left[-\phi_{y}+v^{2}\left ( \frac{1}{v^{4}}+\frac{1}{v^{3}}\right){\phi}^2\right],\ ] ] and then @xmath172 -2{\phi}{\phi_x}\frac{v^1}{v^4 } -2{\phi}^{3}\frac{({v^1})^2}{v^{3}v^{4}}+ 2{\phi}^{3}\left(\frac{v^{1}}{v^{4}}\right)^{2}\right\ } , \\
\tilde{\phi}_{yy } & = & \frac{v^3}{v^4}\left\{-\phi_{yy } + 2{\phi_y^2}{\phi}^{-1 } + \phi^2\left(\frac{1}{v^{4}}+\frac{1}{v^{3}}\right ) \left[-{\phi}v^{0}+\phi_{x}{\phi}^{-1}v^{1 } -\phi_{y}{\phi}^{-1}v^{2 } \right.\right .
\\ & & \left.\left .
{ } + { \phi}v^{3 } + \left(\frac{s}{2}{\phi}^{-3 } -m{\phi}^{-1}\right)v^{4 } \right ] -2{\phi}{\phi_y}\frac{v^2}{v^4 } -2{\phi}^{3}\frac{({v^2})^2}{v^{3}v^{4}}+ 2{\phi}^{3}\left(\frac{v^{2}}{v^{4}}\right)^{2}\right\}.\end{aligned}\ ] ] we now use the constraint @xmath173 to obtain @xmath174 from this equation and the constraint @xmath175 the result follows .
let @xmath25 be a potential with character @xmath176 , then the _ complementary potential _
@xmath177 is defined by @xmath178 .
this is a new potential with the same character of @xmath25 .
note that @xmath177 can be obtained as the bcklund transform of @xmath25 with respect to the system of transforming functions @xmath179 consider the trivial solution @xmath180 of ( [ solitoneq ] ) with @xmath38 . in this case
the @xmath157-form @xmath181 corresponding to the lax pair is given by @xmath182 now set @xmath183 by solving a system of first order linear differential equations with constant coefficients , the normal framing @xmath89 is computed to be @xmath184 in order to find the solutions of ( [ darbouxsys ] ) satisfying the constraint @xmath185 , we may assume that @xmath186 .
note that the pointing vector of the normal framing @xmath89 is the space - like vector @xmath187 .
thus , the transforming functions of the @xmath92-system are given by @xmath188 where @xmath189 is defined by @xmath190 and where @xmath191 is as in ( [ innerproduct ] ) .
it follows that @xmath192 where @xmath193 from this we obtain the following formula for the one - soliton solutions @xmath194
in this section we briefly describe the geometric transformation of special isothermic surfaces corresponding to the bcklund transformation . a _ curved flat framing _ is a smooth map @xmath195 such that @xmath196 takes the form @xmath197 where @xmath5 , @xmath6 , @xmath7 are smooth functions with @xmath198 , for each @xmath199 , and @xmath200 is a constant referred to as the _ spectral parameter _ of the curved flat . for short , the connection form @xmath196 will be denoted by @xmath201 .
the functions @xmath5 , @xmath6 and @xmath7 satisfy the _ isothermic gauss - codazzi system _
@xmath202 note that if @xmath195 is a curved flat framing , then @xmath203:u\to \mathcal{m}$ ] and @xmath204:u\to \mathcal{m}$ ] are isothermic immersions with calapso potentials @xmath205 respectively .
we say that @xmath206 is a curved flat framing along @xmath8 and call @xmath25 the calapso potential of @xmath206 .
the _ dual frame field _ is defined to be @xmath207 , where @xmath208 is given by @xmath209 in particular , @xmath210 is a curved flat framing along @xmath211 .
the mapping @xmath212 defines a sphere congruence whose envelopes are @xmath8 and @xmath211 .
moreover , the correspondence induced by @xmath213 preserves the curvature lines and is conformal .
this amounts to saying that @xmath213 is a darboux congruence .
accordingly , @xmath211 is called a _ darboux transform _ of @xmath8 .
any darboux transform of @xmath8 arises in this way . has been proved in @xcite . for the interpretation of the darboux
transforms as dressing transformations of loop groups we refer to burstall @xcite ] the first envelope @xmath214 of a curved flat framing @xmath206 is a special isothermic immersion if and only if , in addition to ( [ gmsystem ] ) , the triple @xmath215 satisfies an equation of the form @xmath216 where @xmath206 , @xmath217 , @xmath218 are real constants and @xmath13 being @xmath14 and @xmath219 .
we then say that @xmath206 is a _ special curved flat framing _ of type @xmath0 .
notice that both the envelopes of the curved flat are special isothermic immersions if and only if @xmath220 . in this case
we shall say that @xmath211 is a _ special darboux transform _ of @xmath8 .
we now have : let @xmath214 be a special isothermic immersion of type @xmath0 with deformation parameter @xmath20 and calapso potential @xmath25 , and let @xmath221 be the bryant s central frame field along @xmath8 .
next , let @xmath222 be any solution of the @xmath223-linear system @xmath224 satisfying the constraint @xmath225 . then @xmath226\ ] ] defines a special isothermic immersion which is a special darboux transform of @xmath8 such that @xmath227 the connection form @xmath137 of the framing @xmath228 @xmath229+g^{+}(v)^{-1}\beta(m)g^{+}(v)\ ] ] takes the form @xmath230 , with @xmath231 it then follows that @xmath206 is a special curved flat framing of type @xmath232 , where @xmath75 denotes the character of @xmath25 .
the first envelope @xmath233 $ ] represents the original immersion @xmath8 and the second envelope @xmath234 $ ] represents the darboux transform @xmath211 .
if @xmath225 , then @xmath211 is a special isothermic immersion and @xmath235 .
@xmath211 has deformation parameter @xmath20 and the normal frame field @xmath236 along @xmath211 is given by @xmath237 where @xmath238 is the smooth map defined by @xmath239 special isothermic maps with calapso potential @xmath37 are given ( up to the action of the conformal group ) by the following formulae ( where @xmath20 is the deformation parameter of the family ) * if @xmath186 : + @xmath240 * if @xmath241 : + @xmath242 * if @xmath243 : + @xmath244 the surfaces @xmath245 parametrized by the maps @xmath246 are the dupin cyclides .
it is a classical result that dupin cyclides are conformally equivalent to either a circular cone , a circular cylinder , or a torus of revolution @xcite .
the normal frame field along @xmath247 is computed to be : * if @xmath186 : + @xmath248 * if @xmath241 : + @xmath249 * if @xmath243 : + @xmath250 according to the above discussion , the special darboux transforms of @xmath247 are given by @xmath251 , \label{darbouxtransformcyclide}\ ] ] where @xmath252 is the solution of the @xmath223-linear system for @xmath37 with initial condition @xmath253 given by ( [ transformingfunctions ] ) .
thus , @xmath254 is a special isothermic immersion with calapso potential @xmath255 .
let @xmath256 be a wave potential with character @xmath75 , and let @xmath257 and @xmath258 be two systems of @xmath259 and @xmath260-transforming functions , respectively , @xmath261 .
let @xmath262 and @xmath263 be the corresponding bcklund transforms .
then , @xmath264.\label{sovrapposizione}\ ] ] is a wave potential with character @xmath75 .
let @xmath221 be the normal frame with potential @xmath25 and spectral parameter @xmath20 , @xmath265 .
consider the curved flat framing @xmath266 and its dual framing @xmath267 ; then @xmath210 is a curved flat framing and the corresponding normal frame @xmath268 is computed to be @xmath269 where @xmath238 is the smooth map defined as in ( [ tequation ] ) .
next , consider the map @xmath270 given by @xmath271 then @xmath272 is a curved flat framing with spectral parameter @xmath273 and calapso potential @xmath274 .
further , let @xmath275 be defined by @xmath276 so that @xmath277[g^{+}(t^{-1})g^{+}(y)]=\bar{b}(m)g^{+}(l).\ ] ] since @xmath278 is a curved flat framing with calapso potential @xmath279 and spectral parameter @xmath273 , then @xmath280 is a solution of the @xmath281-system with potential @xmath279 . combining ( [ tequation ] ) , ( [ yequation ] ) , ( [ lequation ] ) and using the constraints @xmath282 , it is a computational matter to check that also @xmath280 satisfies the constraint @xmath283 this implies that @xmath280 is a system of @xmath281-transforming functions for the potential @xmath274 .
in particular , the bcklund transform @xmath284 is a new solution of the differential equation ( [ solitoneq ] ) . on the other hand , @xmath284 is computed to be @xmath285 from which follows that @xmath286.\ ] ] the wave potential @xmath287 can be realized as the calapso potential of the special isothermic immersion @xmath288,\label{foursurface}\ ] ] which can be explicitly computed from the normal frame @xmath89 and the two sets @xmath28 and @xmath35 of transforming functions .
[ * two - solitons and the corresponding isothermic surfaces * ] two - soliton solutions can be computed by means of ( [ transformingfunctions ] ) and ( [ sovrapposizione ] ) .
we then obtain a six - parameter family of wave potentials given by @xmath289 where @xmath28 and @xmath35 are the @xmath259 and @xmath260-transforming functions corresponding to the initial conditions @xmath253 and @xmath290 .
special isothermic immersions with spectral parameter @xmath20 and calapso potentials @xmath25 can be constructed by using ( [ foursurface ] ) , ( [ transformingfunctions ] ) and the explicit formulae for the central frame fields @xmath89 of dupin cyclides .
d. carf , e. musso , _ t - transformations of willmore isothermic surfaces _ ,
rend . sem . mat .
messina ser.ii , suppl .
( 2000 ) , 257266 .
j. cieliski , p. goldstein , a. sym , _ isothermic surfaces in @xmath292 as soliton surfaces _ , physics letters * a 205*(1995 ) , pp .
3743 .
e. musso , isothermic surfaces in euclidean space , in _ proc .
workshop on recent topics in differential geometry _
cordero & e. garca - ro , edd .
geometra y topologa , univ .
santiago de compostela , # 89 , santiago de compostela , ( 1998 ) , pp . | we establish a correspondence between darboux s special isothermic surfaces of type @xmath0 and the solutions of the second order p.d.e .
@xmath1 .
we then use the classical darboux transformation for isothermic surfaces to construct a bcklund transformation for this equation and prove a superposition formula for its solutions . as an application
we discuss 1 and 2-soliton solutions and the corresponding surfaces . |
the origin of short grbs has long been the subject of theoretical speculation .
a long - standing prediction is that short grbs might originate from ns+ns / bh mergers ( e.g. , @xcite , and references therein ) . in the widely favored scenario
a ns+ns / bh system gives rise to a bh girded by a dense torus of gas from the disrupted ns companion .
the bh accretes mass at rates up to many solar masses per second , releasing huge amounts of gravitational binding energy mostly in neutrinos , gravitational waves , and through magnetohydrodynamic processes .
hydrodynamic simulations ( @xcite @xcite ; ajm05 ) show that thermal energy deposition around post - merger bh - torus systems , e.g. , by neutrino - antineutrino annihilation , can drive collimated , ultrarelativistic outflows with the high lorentz factors , internal variability and internal shocks that are deemed necessary to explain grbs with the fireball model ( e.g. , @xcite ) .
the collimation of observed short grbs is rather uncertain and a matter of ongoing studies .
evidence for collimation was claimed in case of grb 050709 ( @xcite ) , for which a jet semi - opening angle of @xmath3 was inferred .
although more observations are needed to confirm the collimation of short grbs , this first indication seems to agree with the hydrodynamic simulations of ajm05 .
if the majority of short bursts has opening angles of that size , it would mean that only about 1% of all bursts point to earth .
only in the past year have a few short grbs been localised .
their properties are recapped in table 1 .
two of these short grbs are hosted by elliptical galaxies , and also the host galaxy of grb 051221 appears to have a relatively evolved population of stars , despite its higher star - formation rate .
this adds confidence to the idea that the progenitors of short grbs are old systems .
in addition , grb 050509b is in the outskirts of its potential host , in agreement with expectations that coalescing compact binaries can travel large distances away from their birth sites during their long gravitational - wave driven inspiral @xcite .
moreover , a possible sn association has been ruled out for grb 050509b @xcite and for grb 050709 @xcite .
three out of five short grbs were detected at similar redshifts ( @xmath4 ) , but their intrinsic isotropic - equivalent energy release in gamma rays , @xmath5 , spans a factor of @xmath6100 .
the observation of intrinsically less energetic short grbs at higher redshifts may be selected against . in this paper
we use the jet models of ajm05 to discuss theoretical predictions for the observable properties of short grbs as a function of viewing angle ( sect .
[ sec : jets ] ) . in sect .
[ sec : observability ] we investigate the consequences of these models for the probability of observing short bursts with different isotropic equivalent energies and lorentz factors , including the selection effects due to cosmological redshift . a summary and conclusions follow in sect .
[ sec : conclusions ] .
by means of relativistic hydrodynamics simulations ajm05 studied the formation of ultrarelativistic outflows from bh - torus systems .
the bh and torus masses and the deposition of thermal energy around the bh were chosen as expected from ns+ns / bh merger models . for different assumptions about the environment density and for varied geometry and time - dependence of the energy deposition rate , ajm05 followed the acceleration , collimation , and propagation of the ultrarelativistic outflows for an evolution time of 0.5 seconds .
if the deposition rate of thermal energy per solid angle around the bh was sufficiently large , ultrarelativistic jets were launched along the rotation axis .
although these simulations are still far from taking into account all potentially relevant physics and did not self - consistently track the viscosity - driven torus evolution and its neutrino emission , they nevertheless provide useful insight into the conditions and properties of mass outflows from post - merger bh - torus systems . the fact that an ultrarelativistic jet is launched from a `` naked '' bh - torus system has important consequences for its properties . while in the case of collapsars
as sources of long grbs the ultrarelativistic outflow originates from bh - torus systems at the center of a massive star , the polar jets in ns+ns / bh mergers do not have to plough through many solar masses of overlying stellar matter .
since the acceleration is not damped by swept - up matter , the jets very quickly reach lorentz factors of a few .
the acceleration is driven by the enormous radiation pressure of the pair - photon fireball produced by the energy release around the bh .
collimation of the baryon - poor jets is provided by the much denser torus walls which gird the evacuated polar regions of the bh . as the jets propagate into the extremely low - density environment
, they continue to accelerate , reaching maximum lorentz factors of a few 100 within the simulated evolution time . as a consequence of the interaction with the torus matter at the jet basis and the subsequent free expansion
, the collimated ultrarelativistic outflows possess flat core profiles with only little variation of radially - averaged specific properties .
these cores are bounded at their lateral edges by very steep gradients , which are not smoothed by the prolonged entrainment of baryons as in case of long grb jets .
the rapid decrease of the isotropic - equivalent energy as a function of @xmath7 implies , among other effects , that the probability of observing the lateral , lower lorentz factor wings must be expected to be significantly reduced .
figure [ fig : jetproperties ] provides an overview of the hydrodynamics results for the sample of `` type - b '' models discussed by ajm05 . in these simulations
the bh - torus system was assumed to be surrounded by a very rarified medium .
the models differ in the adopted rate of thermal energy deposition per solid angle and in the prescribed time dependence .
the parameters were chosen such that the neutrino - antineutrino annihilation as calculated from ns+ns / bh merger simulations , and in particular for the post - merger accretion of a bh @xcite , was qualitatively and quantitatively reproduced .
because of large variations of the torus masses expected from compact object mergers , ajm05 explored a variety of energy deposition properties , approximately covering the range of predictions from the above merger simulations ( cf .
table 1 in ajm05 ) .
the corresponding differences in the assumed energy deposition around the bh - torus systems account for the variation of the on - axis values of the _ isotropic - equivalent kinetic plus internal energies _ , @xmath8 , of jet matter with estimated terminal lorentz factors @xmath9 ( fig .
[ fig : jetproperties ] , left ) .
@xmath10 is estimated by assuming that 50% of the total specific energy will ultimately be converted to kinetic energy at large radii is beyond the scope of the hydrodynamic models of ajm05 , because the simulations do not include physics that plays a role during the later stages of the jet acceleration . ] .
the models span a range of @xmath11 in @xmath12 , from some @xmath13erg to several 10@xmath14erg . depending on the efficiency of energy conversion to @xmath15-ray emission
, the measurable @xmath15-ray energy ( @xmath5 ) may be roughly a factor of 10 lower than @xmath8 ( e.g. , @xcite , and references therein ) .
all models show a flat core and steep lateral wings in the radial average of @xmath16 ( fig .
[ fig : jetproperties ] , right ) , but models b07 and b08 have a somewhat more shallow decline of @xmath12 because a gradual decrease of the energy deposition rate was assumed after an initial , burst - like phase .
this caused a continuous decrease of the jet opening angle , thus softening the wings of the jet profile .
it is currently not clear whether the jet - driving energy release of the bh - torus system follows such a burst / slow - decay behavior or whether it is powerful and roughly constant over a longer period of time before it ceases more abruptly , as assumed in the other models ( b1b6 ) .
figure [ fig : jetproperties2 ] provides some observationally relevant quantities deduced from the hydrodynamics results .
the profiles of @xmath8 given in the left panel there represent the isotropic - equivalent values of energy which is potentially radiated from the outflow in different directions .
these profiles are calculated from the jet energies of fig .
[ fig : jetproperties ] , taking into account radiation contributions coming from regions outside of the line of sight . in order to compute these contributions ,
let us consider an amount of energy @xmath17 that is radiated into the solid angle @xmath18 at angle @xmath19 relative to the direction of motion ( primed quantities are measured in the local comoving frame ) .
the transformation of the energy per solid angle between the comoving frame and the laboratory frame ( e.g. , a frame at rest with respect to the central bh ) is ( @xcite ) @xmath20 where @xmath21 is the velocity of the comoving frame as measured in the laboratory frame .
comparing the values of @xmath22 at two different angles , @xmath23 and @xmath24 , one finds that the energy emitted per unit of solid angle around an angle @xmath25 can be expressed in terms of the energy radiated along the direction of propagation ( @xmath24 ) as @xmath26 if the radiation field is isotropic in the comoving frame .
the emission from the jet in a certain observer direction @xmath25 can now be obtained as the sum of the contributions from all matter moving in different directions with velocities @xmath27 at angles @xmath28 relative to the jet axis , yielding @xmath29 since the profiles of @xmath30 displayed in fig .
[ fig : jetproperties ] are related to the energy radiated per unit solid angle into different directions @xmath31 , the integration given in eq .
( [ eq : ecorr ] ) can be directly applied to these profiles in order to obtain the isotropic equivalent energy distributions corrected for off - line - of - sight contributions .
the results are plotted in fig .
[ fig : jetproperties2 ] . from the latter figure it is obvious that energy release with @xmath32erg is confined to semi - opening angles between about 10@xmath33 and 20@xmath33 , corresponding to a collimation ( `` beaming '' ) factor of the two polar jets of @xmath34 between 1.5% and 6% .
the relative observability of a short burst from a model of our set at different polar angles @xmath7 is shown in fig .
[ fig : jetproperties2 ] , middle panel .
we plotted there the quantity @xmath35 where as a rough estimate for the detectability of a burst we assumed @xmath36 ( using for @xmath8 the data from the left panel in fig .
[ fig : jetproperties2 ] ) .
the probability distribution confirms the steep lateral edges visible in the other quantities for models b01b06 and the slightly softer wings in models b07 and b08 , which correlate with a sharp drop of the average lorentz factor .
the probability @xmath37 peaks at angles between 4@xmath33 and @xmath611@xmath33 , which is a somewhat smaller range than that associated with @xmath38erg .
morever , @xmath37 nicely demonstrates the widening of the lateral visibility of a grb with increasing energy deposition per solid angle around the bh ( cf . the sequence of models b02 , b04 , b06 , b01 , and b05 ) .
the hydrodynamic jet models of ajm05 track the formation of ultrarelativistic grb - viable outflow .
it turns out that the `` shells '' ejected by the central engine , i.e. the bh - torus system , accelerate much faster in the leading part of the outflow than the shells in its lagging part .
the rear shells need therefore a longer time to reach velocities @xmath39 .
this different acceleration at early and late times of the relativistic wind ejection leads to a stretching of the overall radial length of the outflow , @xmath40 , relative to the on - time @xmath41 of the central engine times the speed of light @xmath42 , @xmath43 ( see fig .
[ fig : stretching ] ) .
this stretching has the important consequence that the overall observable duration of the grb ( in the source frame ) , @xmath44 , may be a factor of ten or more longer than the activity time of the central energy source ( fig .
[ fig : stretching ] ) , even when the grb is produced by internal shocks .
this is only seemingly in conflict with the canonical fireball picture as discussed , e.g. , by @xcite , who draw the conclusion that `` internal shocks continue as long as the source is active , thus the overall grb duration @xmath45 reflects the time that the inner engine is active '' .
it should in fact be noted that grb light curve calculations in these shell collision models are based on the assumption that the shells move with _ constant lorentz factors and with a velocity very close to the speed of light _ , which means that the expansion of the shells is considered only _ after the preceding acceleration away from the central engine_. therefore the `` shell emission time from the inner engine '' @xcite or `` ejection time '' @xcite or `` activity time of the inner engine '' @xcite referred to in these works corresponds to the overall length the fireball has attained _ in the saturated stage _ after the initial acceleration .
this time can only be identified with the real activity time or on - time of the central engine , whose energy output launches and drives the initial shell ejection , if the shells accelerate to nearly the speed of light within a negligibly small period of time . only in this case
the claim of equal emission and source activity times applies and can be traced back to the fact that a photon radiated from a fireball shell is observed almost simultaneously with a ( hypothetical ) photon launched at the central engine together with the radiating shell @xcite .
the hydrodynamic jet simulations , however , show that the acceleration time , in particular for the shells ejected later , is not negligible ( fig .
[ fig : stretching ] ) . taking the terminal length @xmath46 of the ultrarelativistic outflow in direction @xmath7 as a very crude measure of the overall observed duration @xmath47 of a burst as the radial extension of those parts of the outflow which reach terminal lorentz factors @xmath9 , where @xmath10 of a mass element in the outflow is again calculated by assuming that 50% of its total energy will finally be converted to kinetic energy . ] , we have plotted this duration relative to the on - axis value in the right - hand panel of fig . [ fig : jetproperties2 ] .
the off - axis variations of the possible grb durations are sizable .
bursts which are seen more than 5@xmath4810@xmath48 off - axis have only 30%60% of the on - axis duration .
the outflow stretching can easily be the dominant effect determining the duration of short grbs , but it is not possible to assess its relevance for long bursts without hydrodynamical modeling .
this is due to the more complex and largely different dynamics of the collimated relativistic outflow caused by the presence of the massive star in the collapsar model .
predictions of the absolute duration of the grb emission based on jet simulations where the period of energy release was just a parameter are hard to make .
self - consistent torus evolution models , including the energy production mechanism and the feedback from jet formation , will be needed to provide reliable numbers for the duration of the source activity and the differential acceleration of the ejecta shells at early and late times . moreover
, the jet structure at 0.5@xmath49s after the jet creation , at which time ajm05 had to stop their calculations , may not be representative for the situation at the time when the grb is produced .
we can employ our set of jet models to attempt to predict the observable short - grb distributions versus redshift @xmath0 , fluence , isotropic equivalent energy output in @xmath15-rays , @xmath5 , and lorentz factor @xmath16 of the flow that produces the bulk of the measured @xmath15-radiation .
we discuss these aspects in view of the observed bursts of table 1 .
our methodology is different from that of other groups , who used observational data of grb redshift , luminosity , and peak flux distributions to derive constraints on the intrinsic properties ( event rates or lifetimes , jet collimation , luminosities ) as functions of redshift ( e.g. , @xcite ) .
we instead refer to the sample of jet models from ajm05 to define the intrinsic distribution of short grb energies as a function of viewing angle for the individual events .
this approach suffers from the huge uncertainties of the theoretical models ( see below ) and naturally can have only a tentative character . in order to perform the analysis
, we assume that our models have equal probability and are representative of the intrinsic variability of the short grb source population at all redshifts .
this is a crucial assumption and it can certainly be questioned , but currently it can hardly be replaced by more realistic alternatives , because the uncertainties concerning the link between ns+ns / bh binary parameters and grb properties are still large . in order to test the sensitivity of our analysis to a variation of the comoving short grb rate density as function of redshift , @xmath50 , we consider three cases , namely an intrinsic grb rate which ( i ) is constant with comoving volume , ( ii ) follows the star formation rate sf2@xmath51sfr of eq .
( 4 ) in @xcite , or ( iii ) varies according to the binary neutron star merger rate , case sf2+delay of eq .
( 6 ) in @xcite . based on these prescriptions , we produce the expected observable distributions of grb properties by monte carlo sampling , randomly drawing grb energies and lorentz factors as functions of viewing angle @xmath7 from our set of grb - jet models , using the results shown in fig . [
fig : jetproperties2 ] and fig .
[ fig : jetproperties ] , respectively .
we estimate the isotropic - equivalent grb energy output , @xmath52 , by reducing the isotropic equivalent kinetic plus internal energy of the outflow , @xmath53 of fig .
[ fig : jetproperties2 ] , by a factor of 10 .
this is supposed to account for the limited efficiency of energy conversion to @xmath15-rays and the fact that only a fraction of the @xmath15-emission occurs in the energy band of a measurement - radiation , and model - dependent spectral properties of the @xmath15-ray emission are disregarded as well . ] .
our set of models can be considered to define a co - moving space distribution function @xmath54 , which is assumed to be normalized to unity and whose integral over @xmath16 yields another normalized distribution function , @xmath55 .
the random angle between the jet axis and the line of sight is denoted by @xmath7 .
formally , the expected redshift distribution of measured bursts can be computed from the intrinsic distribution as @xmath56 when @xmath57 represents the observed rate of events in the redshift interval @xmath58 ( in the following we will only consider normalized distributions and the absolute value of the rate will be of no concern ) . in eq .
( [ eq : eventsz ] ) , @xmath59 , @xmath60^{-1}$ ] is the comoving volume element , and @xmath61 accounts for cosmological time dilation .
the cosmological parameters used in our study are @xmath62km@xmath49s@xmath63mpc@xmath63 , @xmath64 , @xmath65 , and @xmath66 . in eq .
( [ eq : eventsz ] ) as well as in eqs .
( [ eq : eventsg])([eq : eventse ] ) below , an instrument specific detection probability that depends on the photon flux and thus on the burst luminosity , spectrum , and redshift , is ignored ( we have no information on this quantity in terms of the fluence ) .
@xmath67 is the isotropic equivalent energy corresponding to the minimum fluence @xmath68 that can be measured by the detector , and @xmath69 corresponds to the maximum energy release of the models of our sample in a certain direction @xmath7 .
the fluence at the instrument is computed from the intrinsic energy output @xmath52 ( in the reference frame of the source ) and from the luminosity distance @xmath70 as @xmath71 $ ] .
the predicted distribution of lorentz factors of the observed bursts can be written as @xmath72 where @xmath73 is the maximum redshift at which grbs with isotropic equivalent energy @xmath52 produce a fluence above the lower detection bound @xmath68 .
similarly , the fluence distribution is given by @xmath74 with @xmath75 , when @xmath76 and @xmath77 are the isotropic equivalent energies of @xmath15-radiation that account for fluences @xmath78 and @xmath79 in the frequency window of the detector for a grb at redshift @xmath0 .
the distribution versus @xmath5 can be represented by @xmath80 we perform our investigation with two different values of the lower fluence cutoff for detection ( i.e. , the detection threshold is assumed to be a step function ) .
for one simulation we use @xmath81erg@xmath49@xmath82 , which corresponds to the fluence of the very weak grb 050509b and thus can be considered as a lower bound of the sensitivity of swift . in a second simulation we use a cutoff value of @xmath83erg@xmath49@xmath82 , which we derive from the limiting flux , @xmath84 $ ] , of 1@xmath49photon@xmath49@xmath82s@xmath63 adopted for the batse instrument by @xcite , making the assumption that @xmath85kev@xmath49s is a representative value for the product of photon detection time interval and energy ( @xmath86 is the intrinsic duration of the burst ) .
this higher fluence threshold seems to be compatible with the sample of ( bright ) batse bursts listed in table 1 of @xcite .
we note that nakar et al .
( @xcite ) employ a detection threshold of 1@xmath49photon@xmath49@xmath82s@xmath63 also for swift . a higher value for @xmath68 may further be motivated by the fact that the photon flux decreases with increasing redshift more rapidly than the fluence by an additional factor @xmath61 . therefore a low fluence threshold for detection may overestimate the number of high-@xmath0 events observed .
to obtain normalized distributions , @xmath87 , we do not evaluate the integrals in eqs .
( [ eq : eventsz])([eq : eventse ] ) directly , but perform monte carlo sampling of a large number of grb events at different redshifts , with random orientations and with properties drawn randomly from the model data plotted in figs .
[ fig : jetproperties ] and [ fig : jetproperties2 ] , thus constructing the discrete distribution functions @xmath88 and @xmath89 which describe the grb properties according to our set of jet simulations .
the sampled events with @xmath90 are then collected into bins to build up the normalized probability distributions versus redshift , fluence , @xmath5 , and @xmath16 shown in fig .
[ fig : grbobservability ] .
we point out that the lorentz factors are based on those of fig .
[ fig : jetproperties ] and therefore reflect the situation only 0.5@xmath49s after the ejection of relativistically expanding matter . because of subsequent acceleration and conversion of internal to kinetic energy ( which we are unable to trace in the hydrodynamic simulations )
, the terminal values @xmath10 must be expected to be larger by a factor of 23 .
our predictions for the observable distributions vs. redshift , fluence , energy , and lorentz factor for the two chosen values of the fluence cutoff are displayed in fig .
[ fig : grbobservability ] . the corresponding locations of four of the five bursts listed in table 1
are also indicated .
the very bright short grb 051221 is outside of the energy scale of the third panel , hence we omitted it from fig .
[ fig : grbobservability ] .
this burst is too energetic to be accounted for by the range of @xmath15-burst energies considered in our set of jet models , in which we assumed that 10% of the ejecta energy can be converted to radiation in the frequency band of the measurement .
but if this fraction were significantly higher the observations of grb 051221 indeed suggest a total efficiency of 6070% ( @xcite ) then the output of @xmath15-energy of this burst would be within the reach of our models for near - axis observation ( cf . fig .
[ fig : jetproperties ] ) .
our jet models invoked energy deposition rates as obtained for the annihilation of neutrino - antineutrino pairs in ns+ns / bh merger and post - merger accretion simulations ( for details and references to original work , see ajm05 ) . we therefore conclude that grb 051221 , if it is indeed located at a redshift of 0.546 , does not make a strong case for different underlying energy extraction mechanisms ( e.g. , magnetohydrodynamic processes ) and/or different progenitors as speculated by @xcite .
it is obvious that the choice of the lower fluence cutoff has a big impact on the redshift distribution ( left panels ) . for a detection threshold of @xmath91erg@xmath49@xmath82 and our employed comoving grb rate densities ,
the majority of bursts should be seen at redshifts @xmath2 , while with the higher cutoff we predict that bursts with @xmath92 would not be detected .
clearly , the observations of the recent bursts with redshifts between 0.16 and 0.72 are more compatible with the latter case . at the request of one referee
, we performed a kolmogorov - smirnov ( ks ) test ( not including the very energetic grb 051221 ) , which gives significance levels of 5% , 8% , and 20% for the consistency of the observations with the theoretical distributions for the three tested grb rate densities sf2@xmath51sfr , sf2+delay , and `` constant '' , respectively ( the ks probabilities in case of the lower cutoff value are 0.1% , 0.4% and 2% ) .
better agreement is therefore found when the simulation predicts a larger number of nearby events .
highest preference is attributed to the constant grb rate density , in which the relative fraction of visible bursts at @xmath93 is largest .
this is in qualitative agreement with the recent analysis by nakar et al .
( @xcite ; see fig . 5 there ) , who found that the well - localized bursts imply a high local rate of short - hard grbs events , and that long lifetimes of the progenitor systems are favored .
given the small number of events , the results clearly will change significantly with every new observation ( see @xcite ) .
the fluence distributions show the expected increase towards faint events , and the distribution of @xmath5 reveals a strong bias towards events with higher isotropic - equivalent energies , corresponding to near - axis observation .
such events are much more likely to yield a fluence above the threshold values at large redshifts and are also preferred because of the steepness of the jet wings , which reduces the probability of seeing events from those wings .
grb 050509b is an exceptionally weak event , which is well separated from the other observed bursts , and which the theoretical distributions do not account for .
it thus distorts the ks measures for the fluence and energy distributions , which otherwise signal reasonably good compatibility between calculations and observations , in particular for the higher fluence threshold .
grb 050509b is most probably _ not _ an off - axis observation , but requires either an event with lower energy output than for any of the jet models in our sample , or an efficiency of the energy conversion to @xmath15-rays lower than assumed here .
the latter is disfavored by the observations of @xcite .
these authors also discard grb 050509b as an off - axis event on the basis of the very early decay of the afterglow light curve .
the probability distribution of the lorentz factor reflects the discreteness of our sample of models , because the redshift has no influence on @xmath16 and because the jet wings are extremely steep and the off - axis visibility much reduced ( see fig .
[ fig : jetproperties ] ) .
therefore , the high - energy model b05 with a relatively wide jet clearly sticks out in the @xmath16-distribution .
a continuous set of models would enhance and broaden this peak , and the `` underwood '' of events extending to low values of @xmath16 would even be more reduced if terminal lorentz factors @xmath10 instead of @xmath16 were plotted .
we hypothesize that this clear bias towards observing high-@xmath16 events may be connected with the lack of short - soft grbs in the duration - hardness diagram .
short bursts are on average harder than long bursts , and short - soft bursts are rare or missing , because short grbs develop jets with higher lorentz factors and very steep edges .
the underlying reason behind this effect is the fact that jets from post - merger bh - torus systems do not have to make their way out of a massive star through layers of dense stellar material , which can damp the outflow acceleration by mass entrainment . in the commonly used internal shock scenario for the grb emission
the peak energy of the ( synchrotron ) spectrum , @xmath94 , obeys the relation @xmath95 with @xmath96 ( e.g. , @xcite ) . in view of our jet models
this would imply that short bursts might have lower @xmath94 but nevertheless be harder than long ones because of a steeper increase of the @xmath97 spectrum below the peak .
the results of ghirlanda et al . ( @xcite ) seem to confirm this .
their spectral analysis for a sample of short bright grbs detected by batse in comparison with the spectral properties of long bright batse bursts indeed reveals a significantly harder ( i.e. , less negative ) spectral index @xmath98 rather than a larger peak energy @xmath94 for short grbs .
this leads to an increased hardness ratio if @xmath94 is above the higher energy band of the spectral hardness measure , or overlaps with it , which is the case for almost all of the short grbs listed in table 1 of ghirlanda et al .
( @xcite ) .
we point out that the difference between the peak spectral energies of short and long bursts might even be more pronounced than found by ghirlanda et al .
( @xcite ) if long bursts occurred at typically higher redshifts and the intrinsic values of these energies were compared .
however , the discussion by zhang & mszros ( 2002 ) also leaves possibilities how in the internal shock model higher lorentz factors of short grbs may be compatible with _ higher _ peak spectral energies . in case of poynting - flux dominated outflow ,
zhang & mszros ( 2002 ) expect a direct proportionality , @xmath99 ( see their eq . 21 and case ( b ) in their fig .
3 ) , and for a kinetic - energy dominated fireball they find @xmath100 ( see their eq .
17 ) when @xmath101 is the radius where internal shock collisions produce @xmath15-ray emission ( see , e.g. , piran 2005 ) .
although short grbs have higher @xmath16 s compared with long grbs , their variability timescale @xmath102 might be generically much smaller .
this would result in a smaller shock dissipation radius and therefore in a higher @xmath94 , because magnetic fields may be stronger at a smaller radius .
using a set of hydrodynamic simulations of ultrarelativistic jet formation by thermal energy release around bh - torus systems as remnants of ns+ns / bh mergers ( ajm05 ) , we have investigated the off - axis visibility of short grbs produced by such outflows .
the jets are characterized by narrow cores with high isotropic - equivalent energies and very high lorentz factors , which are laterally bounded by steep wings .
these properties are a consequence of the specific conditions in which the jets are considered to be launched , i.e. , their acceleration away from the torus - girded bh , where the energy is released , into an environment of extremely low density .
the observability of short grbs is therefore strongly favored within a cone of semi - opening angle between about 10@xmath48 and 15@xmath48 around the jet axis .
we argued that the observable duration of short grbs both in the external shock model and in the internal shock collision model for grb production can be significantly longer than the activity time of the central engine whose energy release powers the grb - viable ultrarelativistic outflow .
the reason for this claim is the fact that our jet formation simulations showed that during the acceleration phase the fireball experiences significant radial stretching , because hydrodynamic effects cause the acceleration to proceed differently in the leading and lagging parts of the outflow .
therefore the overall radial length @xmath40 of the relativistic ejecta can become a factor of ten or more larger than @xmath103 , when @xmath41 is the on - time of the central engine ( fig .
[ fig : stretching ] ) .
this seems to be in conflict with the canonical internal shock scenario , according to which the observed grb is produced by shell collisions such that the observed light curve reflects the temporal activity and overall duration of the energy release by the `` inner engine '' @xcite .
however , the standard picture applies only if the fireball shells move immediately after their ejection with velocities that are very similar ( but not necessarily identical ) and very close to the speed of light .
this can be traced back to the fact that a photon radiated from a fireball shell is observed almost simultaneously with a ( hypothetical ) photon emitted from the central engine together with the radiating shell @xcite .
if the acceleration is differential , and in particular if the acceleration of shells ejected by the central engine at later times proceeds more slowly , the fireball evolution after creation at the engine is more complex ( fig .
[ fig : stretching ] ) .
its emission pattern then does not directly replicate the activity timescales and the on - time of the central source . in order to investigate the consequences of the reduced off - axis visibility of short grbs produced by our jets
, we made an attempt to predict observable short - hard grb distributions with redshift , fluence , isotropic - equivalent energy , and lorentz factor .
for this purpose we made use of a sample of seven ultrarelativistic jet models ( ajm05 ) , which roughly covers the range of energy release expected from neutrino - antineutrino annihilation in ns+ns / bh binary mergers . employing these models for defining the intrinsic grb properties
, we performed a monte carlo sampling of the expected observed event distributions .
our analysis took into account the off - axis variation of the jet properties as provided by the hydrodynamic simulations ( figs .
[ fig : jetproperties ] and [ fig : jetproperties2 ] ) and referred to standard prescriptions for the intrinsic event rate density as a function of redshift ( constant rate as well as the cases sf2@xmath51sfr and sf2+delay from @xcite ) .
we found that the resulting distributions predict far too many bursts at @xmath2 and therefore show no good agreement with the recent observations when a fluence threshold of @xmath91erg@xmath49@xmath82 as suggested by grb 050509b was adopted ( the ks probability for the observed and predicted data being drawn from the same distribution is less than 1% ) .
improved consistency was achieved when we performed our analysis with a higher fluence threshold of @xmath104erg@xmath49@xmath82 . in this case
the bursts at @xmath105 escape from observation and the ks significance increases to @xmath620% for the constant comoving grb rate density .
this would imply that grb 050509b was a very special and extremely rare case with an exceptionally low fluence , which due to its shortness and proximity still had a sufficiently large photon flux of about 1@xmath49photon s@xmath106@xmath82 and was therefore above the swift detection threshold .
comparison of the results for different comoving grb rate densities revealed that the consistency between theory and observations increases when the rate yields a relatively larger number of events at redshifts @xmath107 rather than an increase for higher values of @xmath0 .
our results agree qualitatively with those of nakar et al .
( @xcite ) , who found that the detection of the current sample of low-@xmath0 short grbs is best compatible with a very low comoving rate density for @xmath2 and a peak at @xmath107 .
the extremely weak grb 050509b is clearly separated in energy from the other observed events and can not be accounted for by off - axis observation of any of the modelled jet outflows .
it seems to require an intrinsically less energetic event , because an efficiency of energy conversion to @xmath15-rays significantly lower than what we used is disfavored by observations @xcite . on the other hand , the very energetic grb 051221 exceeds the isotropic - equivalent grb energies predicted by our jet models , making the assumption that 10% of the ejecta energy can be converted to radiation in the detector frequency band .
if , however , this fraction is as high as suggested by the measurements ( 6070% for the total efficiency in case of grb 051221 ; @xcite ) and grb 051221 happened indeed at a redshift of 0.546 and not farther away , then this strong burst is also within the reach of the most energetic one of our models if observed near - axis .
therefore we do not think that grb 051221 gives strong support to speculations that different progenitors and/or different energy extraction mechanisms than neutrino - antineutrino annihilation are needed to explain the observed large spread in short burst energies and the high energy of grb 051221 in particular @xcite .
because of the steep wings of the jet profiles in both isotropic - equivalent energy and lorentz factor @xmath16 , our analysis reveals a clear bias towards the detection of short grbs with high @xmath16 values .
since higher lorentz factors and steep jet edges may be characteristic features of grb jets that originate from post - merger bh - torus systems , which makes them different from collapsar jets , we propose that they are the reason why short grbs are typically harder than long ones . in the internal shock scenario for grb emission
, this might imply that the peak energy of the synchrotron spectrum is lower ( @xmath108 ) .
short bursts could therefore be harder not because of a higher @xmath109 but because of a steeper increase of their @xmath97 spectra below the peak energy .
this would require the spectral maximum to lie above or inside the high - energy band for the hardness ratio .
such properties were indeed concluded from a spectral analysis of a sample of short bright batse grbs in comparison with long burst spectra ( ghirlanda et al .
alternatively , a higher @xmath16 could still allow for a larger @xmath109 if the variability timescale @xmath102 of short grbs were generically much smaller than that of long grbs .
we point out that our exploration can only be tentative .
both the theoretical models and observational data still involve large uncertainties , which have an impact on our analysis .
not only did we take a very simplistic approach in estimating the @xmath15-ray production of our jet models in the energy band of the detector .
we just assumed that a fraction of 10% of the ejecta energy is converted to @xmath15-emission at the relevant frequencies .
we also did not account in detail for the instrument specific detection probability that depends on the photon flux and thus on the burst luminosity , spectrum , and redshift , but simply assumed that all short grbs above a certain fluence threshold are detected , independent of their peak luminosity and spectral properties .
moreover , the intrinsic short grb rate per unit of comoving time and comoving volume enters the analysis sensitively .
nakar et al .
( @xcite ) found that the well - localized short grbs , which are all at @xmath110 , require a very large number of nearby events .
this implies that the local number density of double neutron star binaries or neutron star - black hole systems is significantly above previous estimates , if short grbs come from such objects .
in fact , it must be dominated by a so far undetected old population of double neutron stars or , alternatively , by observationally unconstrained ns+bh systems .
a caveat of such conclusions and of comparisons to the observations is , of course , the still very small sample of well - localized short grbs .
caution is therefore advisable until the empirical foundation is more solid ( see @xcite ) .
we restricted our analysis to the properties of the prompt emission of short grbs , which may originate from neutrino - annihilation driven hydrodynamic jets , launched by the neutrino energy release of a hyperaccreting bh girded by a massive torus .
of course , one can not exclude that magnetohydrodynamic effects might play a role that deserves more theoretical exploration .
the discovery of x - ray flares following the short - duration , hard pulse of grb 050724 after hundreds of seconds ( barthelmy et al .
2005 ) was interpreted as observational hint for the importance of such magnetic processes ( fan , zhang , & proga 2005 ) .
if these late - time x - ray flares are indeed connected with an extended period of activity or a restart of the central engine long after the prompt emission is over ( burrows et al .
2005 , zhang et al .
2005 , liang et al .
2006 ) , it seems not possible to explain these late outbursts by the accretion - rate dependent neutrino mechanism ( fan , zhang , & proga 2005 ) . a prompt phase of neutrino ( or neutrino plus magnetic ) energy release that creates the short - hard grb may thus be followed by a much more extended period of reduced source activity , in which x - ray flares could be powered by a magnetic mechanism ( fan et al .
2005 ) .
admittedly , our investigation of the implications of the off - axis properties of short grb jets from post - merger bh - torus systems was also based on a very limited set of numerical models ( ajm05 ) .
this theory input of our analysis replaces the intrinsic peak luminosity function based on measured data used in other works ( e.g. , nakar et al .
@xcite , @xcite ) .
the models were chosen to span approximately the possible diversity of the properties of ultrarelativistic outflows from such systems , but certainly represent it only incompletely . in particular , they were calculated with a single setup for the bh - torus system , in which the deposition rate of thermal energy and its spatial distribution ( expressed in terms of the opening angle of the axial cone of the main energy deposition ) was varied over a range guided by the results for neutrino - antineutrino annihilation in independent binary merger and post - merger evolution models .
neither was the torus evolved self - consistently in response to this energy release and to the jet formation , nor were magnetohydrodynamic effects included in the models .
there is still a long way to go before the challenging problem of fully consistent modeling becomes manageable .
this should include all relevant physics and track in full general relativity the history from the last orbits of the pre - merging binary , through the merging phase , to the possible bh formation and the subsequent secular post - merging evolution of the accreting relic black hole .
such simulations will ultimately have to be performed for a carefully selected sample of progenitor systems , which avoids the `` arbitrariness '' of the variations in our current set of models .
knowing that this goal is still far ahead and in view of the fact that the simulations available to us represent the present state - of - the - art of modeling collimated relativistic outflows from post - merger bh - torus systems , we tried to explore how far this information can help us bridging grb engine theory and observations .
@lccccccccccccccc@ + grb@xmath111 & @xmath0 & band & @xmath112 & @xmath113 & fluence & @xmath114 & @xmath115 & @xmath109 & @xmath116 & @xmath117 & gal . &
d@xmath118 & sfr & ref.@xmath119 + & & kev & ms & ms & @xmath91erg@xmath49@xmath82 & @xmath120erg & @xmath120erg@xmath49s@xmath63 & kev & & & & kpc & @xmath121 & + 050509b & 0.225 & 15150 & 40 & 33 &
@xmath122 & 0.011 & 0.81.5 & @xmath123 150 & @xmath124 & 2.34.3 & ell & 40 & @xmath125 & 1,5 + 050709 & 0.16 & 30400 & 70 & 60 & @xmath126 & 0.167 & 4.7 & @xmath127 & @xmath128 & 1.7 & sf & 3.8 & 0.2 & 2,6 + 050724@xmath129 & 0.257 & 15350 & 250 & 200 & @xmath130 & 0.174 & 1.41.7 & @xmath123350 & @xmath131 & 1.61.9 & ell & 2.6 & @xmath132 & 3,7 + & & & 3000 & 2387 & @xmath133 & 0.97 & 0.70.8 & ... & & & & & & + 050813 & 0.722 & 15350 & 600 & 350 & @xmath134 & 1.6 & 7.813 & @xmath123350 & @xmath135 & 1.72.8 & ? & & & 4,8 + 051221 & 0.546 & 202000 & 1400 & 906 & 320@xmath136 & 24 & 26 & @xmath137 & & 1.0 & sf & 0.8 & 1.5 & 9 + + + + + + + + + + + | based on recent models of relativistic jet formation by thermal energy deposition around black hole - torus systems , the relation between the on- and off - axis appearance of short , hard gamma - ray bursts ( grbs ) is discussed in terms of energetics , duration , average lorentz factor , and probability of observation , assuming that the central engines are remnants of binary neutron star ( ns+ns ) or neutron star - black hole ( ns+bh ) mergers . as a consequence of the interaction with the torus matter at the jet basis and the subsequent expansion of the jets into an extremely low - density environment , the collimated ultrarelativistic outflows possess flat core profiles with only little variation of radially - averaged properties , and are bounded by very steep lateral edges .
owing to the rapid decrease of the isotropic - equivalent energy near the jet edges , the probability of observing the lateral , lower lorentz factor wings is significantly reduced and most short grbs should be seen with on - axis - like properties .
taking into account cosmological and viewing angle effects , theoretical predictions are made for the short - grb distributions with redshift @xmath0 , fluence , and isotropic - equivalent energy .
the observational data for short bursts with determined redshifts are found to be compatible with the predictions only if either the intrinsic grb rate density drops rapidly at @xmath1 , or a large number of events at @xmath2 are missed , implying that the subenergetic grb 050509b was an extremely rare low - fluence event with detectable photon flux only because of its proximity and shortness .
it appears unlikely that grb 050509b can be explained as an off - axis event .
the detection of short grbs with small lorentz factors is statistically disfavored , suggesting a possible reason for the absence of soft short bursts in the duration - hardness diagram . |
the hot gas in the igm induces distortions in the spectrum of the cosmic microwave background ( cmb ) through inverse compton scattering .
this effect , known as the thermal sunyaev - zeldovich ( sz ) effect @xcite , is a source of secondary anisotropies in the temperature of the cmb ( see refs . for reviews ) . because the sz effect is proportional to the integrated pressure of the gas
, it is a direct probe of the large scale structure in the low redshift universe .
moreover , it must be carefully subtracted from the primary cmb anisotropies , to allow the high - precision determination of cosmological parameters with the new generation of cmb experiments ( see and references therein ) .
thanks to impressive recent observational progress , the sz effect from clusters of galaxies is now well established @xcite .
the statistics of sz clusters were calculated by a number of authors using the press - schechter ( ps ) formalism ( eg . ) .
recently , atrio - barandera & mcket used this formalism , along with assumptions about cluster profiles to compute the angular power spectrum of the sz anisotropies for the einstein de sitter universe .
a similar calculation was carried out by komatsu & kitayama ( kk99 , hereafter ) , who also studied the effect of the spatial correlation of clusters and cosmological models .
the statistics of sz anisotropies have also been studied using hydrodynamical simulations .
scaramella et al .
@xcite , and more recently da silva et al .
@xcite , have used this approach to construct sz maps and study their statistical properties .
persi et al .
@xcite instead used a semi - analytical method , consisting of computing the sz angular power spectrum by projecting the 3-dimensional power spectrum of the gas pressure on the sky . in this paper
, we follow the approach of persi et al . using moving mesh hydrodynamical ( mmh ) simulations @xcite .
we focus on the angular power spectrum of the sz effect and study its dependence on cosmology .
we compare our results to the press - schechter predictions derived using the methods of kk99 .
we study the redshift dependence of the sz power spectrum , and estimate the contribution of groups and filaments .
we also study the effect of the finite resolution and finite box size of the simulations .
results from projected maps of the sz effect using the same simulations are presented in seljak et al.@xcite .
we study the implications of our results for future and upcoming cmb missions ( see also refs.@xcite ) .
this paper is organized as follows . in [ sz ] , we briefly describe the sz effect and derive expressions for the integrated comptonization parameter and the sz power spectrum . in [ methods ] , we describe our different methods used to compute these quantities : hydrodynamical simulations , the ps formalism , and a simple model with constant bias .
we present our results in [ results ] , and discuss the limitations imposed by the finite resolution and box size of the simulations .
our conclusions are summarized in [ conclusions ] .
the sz effect is produced from the inverse compton scattering of cmb photons @xcite . the resulting change in the ( thermodynamic ) cmb temperature is @xmath7 where @xmath8 is the unperturbed cmb temperature , @xmath0 is the comptonization parameter , and @xmath9 is a spectral function defined in terms of @xmath10
, @xmath11 is the planck constant and @xmath12 is the boltzmann constant . in the nonrelativistic regime ,
the spectral function is given by @xmath13 , which is negative ( positive ) for observation frequencies @xmath14 below ( above ) @xmath15 ghz , for @xmath16 k. in the rayleigh - jeans ( rj ) limit ( @xmath17 ) , @xmath18 .
the comptonization parameter is given by @xmath19 where @xmath20 is the thomson cross - section , @xmath21 , @xmath22 and @xmath23 are the number density , temperature and thermal pressure of the electrons , respectively , and the integral is over the physical line - of - sight distance @xmath24 .
we consider a general frw background cosmology with a scale parameter defined as @xmath25 , where @xmath26 is the scale radius at time @xmath27 and @xmath28 is its present value .
the friedmann equation implies that @xmath29 where @xmath30 , @xmath31 , and @xmath32 are the present total , matter , and vacuum density in units of the critical density @xmath33 . as usual , the hubble constant today is parametrized by @xmath34 km s@xmath35 mpc@xmath35 .
it is related to the present scale radius by @xmath36 , where @xmath37 , 1 , and @xmath38 in a open , flat , and closed cosmology , respectively .
the comoving distance @xmath39 , the conformal time @xmath40 , the light travel time @xmath27 , and the physical distance @xmath41 are then related by @xmath42 . with these conventions , and assuming that the electrons and ions are in thermal equilibrium , equation ( [ eq : y_basic ] ) becomes @xmath43 where @xmath44 is the gas mass density
, @xmath45 is the gas temperature , and @xmath46 is the number of electrons per proton mass . equation ( [ eq : y_basic2 ] ) can be written in the convenient form @xmath47 where @xmath48 is the gas density - weighted temperature , and @xmath49 .
the overbar denotes a spatial average and @xmath50 is the present baryon density parameter .
the constant @xmath51 is given by @xmath52 where the central value for @xmath53 was chosen to correspond to a he fraction by mass of 0.24 , and that for @xmath50 to agree with big bang nucleosynthesis constraints@xcite .
the mean comptonization parameter @xmath54 can be directly measured from the distortion of the cmb spectrum ( see ref . for a review ) , and is given by @xmath55 it can thus be computed directly from the history of the volume - averaged density - weighted temperature @xmath56 .
the gas in groups and filaments is at a temperature of the order of @xmath57 ( or @xmath58 ) , and thus induce a @xmath0-parameter of the order of @xmath2 over a cosmological distance of @xmath59 ( see eq . [ [ eq : y0 ] ] ) .
this is one order of magnitude below the current upper limit of @xmath60 ( 95% cl ) from the cobe / firas instrument @xcite .
the cmb temperature fluctuations produced by the sz effect are quantified by their spherical harmonics coefficients @xmath61 , which are defined by @xmath62 .
the angular power spectrum of the sz effect is then @xmath63 .
they are related to the power spectrum by @xmath64 , where @xmath65 denotes the 2-dimensional dirac - delta function .
the sz temperature variance is then @xmath66 .
since , as we will see , @xmath67 varies slowly in cosmic time scale and since the pressure fluctuations occur on scales much smaller than the horizon scale , we can apply limber s equation in fourier space ( eg .
ref . ) to equation ( [ eq : y_dchi ] ) and obtain , @xmath68 where @xmath69 , @xmath39 , and @xmath70 are the comoving angular diameter distances in an open , flat , and closed cosmology , respectively , and @xmath71 is the 3-dimensional power spectrum of the pressure fluctuations , at a given comoving distance @xmath39 . in general , we define the 3-dimensional power spectrum @xmath72 of a quantity @xmath73 by @xmath74 where @xmath75 , and @xmath76 . with these conventions ,
the variance is @xmath77 . for a flat universe ,
equation ( [ eq : cl ] ) agrees with the expression of persi et al . @xcite .
the sz power spectrum can thus be readily computed from the history of the mean density - weighted temperature @xmath78 and of the pressure power spectrum @xmath71 .
we used the mmh code written by pen@xcite , which was developed by merging concepts from earlier hydrodynamic methods .
grid - based algorithms feature low computational cost and high resolution per grid element , but have difficulties providing the large dynamic range in length scales necessary for cosmological applications . on the other hand , particle - based schemes , such as the smooth particle hydrodynamics ( for a review see ref . )
fix their resolution in mass elements rather than in space and are able to resolve dense regions .
however , due to the development of shear and vorticity , the nearest neighbors of particles change in time and must be determined dynamically at each time step at a large computational cost .
to resolve these problems , several approaches have recently been suggested @xcite .
the mmh code combines the advantages of both the particle and grid - based approaches by deforming a grid mesh along potential flow lines .
it provides a twenty fold increase in resolution over previous cartesian grid eulerian schemes , while maintaining regular grid conditions everywhere @xcite .
the grid is structured in a way that allows the use of high resolution shock capturing tvd schemes ( see for example ref . and references therein . ) at a low computational cost per grid cell .
the code that optimized for parallel processing , which is straightforward due to the regular mesh structure .
the moving mesh provides linear compression factors of about 10 , which correspond to compression factors of about @xmath79 in density .
note that this code does not include the effects of cooling and feedback of the gas .
we ran three simulation with @xmath80 curvilinear cells , corresponding to @xmath81-normalized scdm , @xmath82cdm , and ocdm models .
the simulation parameters are listed in table [ tab : simulations ] .
note that in all cases , the shape parameter for the linear power spectrum was set to @xmath83 @xcite .
the simulation output was saved at @xmath84 and @xmath85 , and was used to compute 3-dimensional statistics . to test the resolution of the simulation , we compared the power spectrum of the dark matter density fluctuations @xmath86 ( defined in eq .
[ [ eq : p_q ] ] with @xmath87 ) from the simulations to that from the peacock & dodds @xcite fitting formula .
the results for the @xmath82cdm are shown on figure [ fig : pk_rhodm ] , and are similar for the other three models .
the simulation power spectrum agrees well with the fitting formula for @xmath88 mpc@xmath35 at all redshifts . for @xmath89 and @xmath90 mpc@xmath35 ,
the simulations are limited by the finite size of the box and the finite resolution , respectively .
we will use these limits below , to study the effect of these limitations on the sz power spectrum .
[ resolution ] it is useful to compare the simulation results with analytic calculations based on the press schechter ( ps ) formalism@xcite .
we compute the angular power spectrum and the mean comptonization parameter , using the methods of kk99 and barbosa et al.@xcite , respectively . for definitiveness
, we adopt the spherical isothermal @xmath91 model with the gaussian - like filter for the gas density distribution in a cluster , @xmath92^{-3\beta/2 } e^{-r^2/\xi r^2},\ ] ] where @xmath26 and @xmath93 are the virial radius and the core radius of a cluster , respectively , and a fudge factor @xmath94 is taken to properly normalize the gas mass enclosed in a cluster@xcite .
we employed a self - similar model for the cluster evolution@xcite .
note that other evolution models yield spectra that differ only at small angular scales ( @xmath95)@xcite . the gas mass fraction of objects is taken to be the cosmological mean , i.e. , @xmath96 .
the volume - averaged density - weighted temperature is given by @xmath97 where @xmath98 is the present mean mass density of the universe , @xmath99 is the ps mass function which gives the comoving number density of collapsed objects of mass @xmath100 at @xmath101 .
@xmath45 is computed by the virial temperature given by @xmath102 where @xmath103 is the mean mass density of a collapsed object at @xmath101 in units of @xmath104@xcite .
while barbosa et al.@xcite used @xmath105 , we adopt @xmath106 according to kk99 .
the limits @xmath107 and @xmath108 should be taken to fit the resolved mass range in the simulation .
the mass enclosed in the spherical top - hat filter with comoving wavenumber @xmath109 is @xmath110 since the @xmath109-range of confidence in the simulation is approximately @xmath111 ( see [ simulations ] ) , equation ( [ eq : mass ] ) gives @xmath112 and @xmath113 .
this mass range is used for calculating the angular power spectrum , the mean comptonization parameter , and the density - weighted temperature .
a more detailed inspection of figure [ fig : pk_rhodm ] reveals that the resolution of the simulations depends on redshift , and involves a power law cutoff in @xmath109 rather than a sharp cutoff .
this must be kept in mind when comparing the two methods ( see [ limitations ] ) .
it is also useful to consider a simple model with constant bias .
the bias @xmath114 of the pressure with respect to the dm density can be defined as @xmath115 and generally depends both on wave number @xmath109 and redshift @xmath101 . in this simple model , we assume that @xmath114 is independent of both @xmath109 and @xmath101 , and
replace the pressure power spectrum @xmath116 in equation ( [ eq : cl ] ) , by @xmath117 , where @xmath118 is evaluated using the peacock & dodds fitting formula @xcite .
this has the advantage of allowing us to extend the contribution to the sz power spectrum to arbitrary ranges of @xmath109 .
this will be used in [ limitations ] to test the effect of finite resolution and finite box size on the sz power spectrum .
figure [ fig : trho ] shows a map of the density - weighted temperature for the @xmath82cdm model projected through one box at @xmath119 . clusters of galaxies are clearly apparent as regions with @xmath120 kev .
the gas in filaments and groups can be seen to stretch between clusters and has temperatures in the range @xmath121 kev .
while these regions have smaller temperatures , they have a relatively large covering factor and can thus contribute considerably to the @xmath0-parameter and to the sz fluctuations .
this can be seen more clearly in figure [ fig : y ] , which shows the corresponding map of the comptonization parameter .
clusters produce @xmath0-parameters greater than @xmath122 , while groups and filaments produce @xmath0-parameters in the range @xmath123 .
note that of the total sz effect on the sky would include contributions for a number of simulation boxes along the line - of - sight .
in such a map , the filamentary structure is less apparent as filaments are averaged out by projection @xcite .
a quantitative analysis of the contribution of groups and filaments to the sz effect is presented in the following sections .
the evolution of the density - weighted temperature for each of the simulations is shown on figure [ fig : t_z ] .
the temperatures at present are listed in table [ tab : results ] and are quite similar .
this is expected since all models were chosen to have similar @xmath81 normalizations .
the evolution is steeper for the scdm , flatter for the ocdm model , and intermediate for the @xmath82cdm model .
this is consistent with the different rate of growth of structure for each model .
also plotted on this figure is the density - weighted temperature derived from the ps formalism ( eq . [
[ eq : pstrho ] ] ) .
the agreement for @xmath124 is good , both for the relative amplitudes and for the shapes of the temperature evolution . at @xmath125
the non - linear mass scale is not sufficiently large compared to the mass resolution of the simulation , so the temperatures are not meaningful in that regime .
this is however not a serious limitation , as these redshifts do not contribute significantly to either the mean comptonization or the sz fluctuations .
the ps temperatures exceed the simulations at low redshift for all cosmological models , since massive ( high temperature ) clusters , which may be missed in the simulations due to the effect of finite box size , dominate there .
the ps temperatures at @xmath119 are listed in table [ tab : results ] .
the parameters of our @xmath82cdm model were chosen to coincide with that for the simulation of cen & ostriker@xcite .
while the slope of our density - weighted temperature agrees approximately with theirs for @xmath126 , the amplitude is significantly different .
they find a final temperature of about 0.9 kev , which is a factor of about 5 larger than ours .
this discrepancy could be due to the fact that their simulation include feedback from star formation , while ours only comprise gravitational forces .
it is however surprising that standard feedback could produce such a large difference .
one can estimate the gravitational binding energy of virialized matter from the cosmic energy equation @xcite and finds a thermal component of fluids to be of order 1/4 kev , consistent with our simulations .
we should note , however , that the high thermal temperatures from feedback may be required for consistency with the x - ray background constraints@xcite .
the reason for this discrepancy is still unknown at present , but should be kept in mind for the interpretation of our results .
the mean comptonization parameter for each simulation was derived using equation ( [ eq : y_bar ] ) and is listed in table [ tab : results ] . in all cases , @xmath54 is well below the upper limit @xmath60 ( 95% cl ) set by the cobe / firas instrument @xcite .
the differential and cumulative redshift dependence of @xmath54 are shown on figure [ fig : y_z ] .
for the three models , most of the mean sz effect is produced at @xmath127 .
the contribution from high redshift is largest for the ocdm and smallest for the scdm model , again in agreement with the relative growth of fluctuations in each model .
the differential and cumulative redshift dependence of @xmath54 derived from the ps formalism are shown on this figure as the thin lines .
the values of @xmath54 from ps are also listed in table [ tab : results ] .
they are higher than that for the simulations by about 25% for the @xmath82cdm and ocdm models , are in close agreement for the scdm model .
the shapes of the differential curves approximately agree , although the ps formalism predicts more contributions from lower redshifts .
this is can be traced to the slightly steeper evolution of the ps temperatures in figure [ fig : t_z ] , and is due to massive nearby clusters . as noted in
[ sz ] , the sz power spectrum can be derived from the history of the temperature @xmath56 and of the pressure power spectrum @xmath128 ( eq . [ [ eq : cl ] ] ) .
the evolution of the pressure power spectrum is shown on figure [ fig : pk_p ] , for the @xmath82cdm simulation .
the amplitude of @xmath128 increases with redshift , while keeping an approximately similar shape .
perhaps more instructive is the evolution of the pressure bias @xmath129 ( eq [ [ eq : bp ] ] ) , which is shown on figure [ fig : bp ] . for @xmath130
, @xmath114 is approximately independent of scale , in the @xmath109-range of confidence ( @xmath131 mpc@xmath35 ; see figure [ fig : pk_rhodm ] ) .
the value of @xmath114 at @xmath119 and @xmath132 mpc@xmath35 is listed in table [ tab : results ] . for @xmath133
, @xmath114 remains approximately constant on large scales , but is larger on small scales .
indeed , at early times , only a small number of small regions have collapsed and are thus sufficiently hot to contribute to the pressure . as a result , the pressure at high-@xmath101 is more strongly biased on small scales .
the sz angular power spectrum derived from integrating the pressure power spectrum along the line of sight ( eq . [ [ eq : cl ] ] ) is shown in figure [ fig : cl ] for each simulation . for comparison ,
the spectrum of primary cmb anisotropies was computed using cmbfast@xcite , and was also plotted on this figure as the solid line .
the sz power spectrum can be seen to be two order of magnitude below the primordial power spectrum below @xmath134 , but comparable to it beyond that .
because of finite resolution and box size , the sz power spectra should be interpreted as lower limits outside of the @xmath41-range of confidence highlighted by thicker lines ( see [ limitations ] ) . the scdm spectrum is lower than that for the @xmath82cdm and ocdm .
this is a consequence of the lower value of @xmath81 for this model .
indeed , kk99 have shown that the sz power spectrum scales as @xmath135 , and is thus very sensitive on this normalization .
this scaling relation also allows us to compare our results to that of the scdm calculation of persi et al.@xcite .
the amplitude of their power spectrum , rescaled to the same value of @xmath81 , is within 20% of ours at @xmath136 , while its shape is similar to ours .
figure [ fig : cl_methods ] presents a comparison of the sz power spectra derived from each of the three methods described in [ methods ] . for both the simulations and the ps formalism , the sz power spectra peak around @xmath137 for the scdm and @xmath82cdm models , and around @xmath138 for the ocdm model . on the other hand , the constant bias models , which do not have a mass or @xmath109 cutoff , peak at @xmath139
this is explained by the fact that this model does not have a mass or @xmath109 cutoff and has therefore more power on small scales . in [ limitations ]
, we will use this comparison to study the effect of finite resolution and box size of the simulations . for @xmath1 , the simulation and ps predictions approximately agree for the scdm and @xmath82cdm models .
on the other hand , for the ocdm model , the ps prediction is a factor of 3 higher than that from the simulations in this range .
this can be traced to the fact that the @xmath82cdm simulation yields a larger pressure bias @xmath114 ( eq . [ eq : bp ] ) at low redshifts than the ocdm simulation . by inspecting the figure corresponding to figure [ fig : pk_rhodm ] for the ocdm model
, we indeed noticed that more power was missing on small scales in this simulation .
this is probably due to the fact that the ocdm simulation was started at a higher redshift ( @xmath140 ) than the other two simulations ( @xmath141 ) .
due to truncation errors in the laplacian and gradient calculations , modes with frequencies close to the nyquist frequency are known to grow much more slowly even in the linear regime .
this effect is reduced if the simulation is started later .
the redshift dependence of the sz power spectrum is shown in figure [ fig : cl_z ] , for the @xmath82cdm model .
most of the sz fluctuations are produced at low redshifts : at @xmath5 , about 50% of the power spectrum is produced at @xmath142 , and about 90% at @xmath143 .
the contribution of the warm gas in groups and filaments can be studied by examining figure [ fig : cl_t ] .
this figure shows the @xmath82cdm power spectrum measured after removing hot regions from the simulation volume , for several cutoff temperatures .
approximately 50% of the sz power spectrum at @xmath5 is produced by gas with @xmath144 . in [ prospects ] , we show that these combined facts give good prospects for the removal ( and the detection ) of sz fluctuations from cmb maps .
the behavior of the power spectrum for @xmath145 , in figures [ fig : cl_t ] and [ fig : cl_z ] , agrees with the results of kk99 who studied the poisson and clustering contributions separately . at low @xmath41 s
, the sz power spectrum is produced primarily by bright ( low - redshift or high - temperature ) objects , i.e. , by massive clusters , and is thus dominated by the poisson term .
however , after subtracting bright clusters from the sz map , the correlation term dominates the poisson term at high redshift . therefore , the sz spectrum on large angular , measured after subtracting bright spots , should trace clustering at high redshift .
this interesting effect will be discussed in details elsewhere .
it is important to assess the effect of the limitations of the simulations on these results .
first , the finite resolution may lower the temperature @xmath146 , since it prevents small scale structures from collapsing . as we saw in
[ ps ] , the resolution limits of the simulations correspond to halo masses of about @xmath147 . according to the ps formalism , the contribution to @xmath146 from halos with masses smaller than this limit is about @xmath148 kev , for @xmath124 in the @xmath82cdm model .
the sz power spectrum at @xmath134 is produced mainly at low redshifts , and is therefore little affected by this effect .
note however , that @xmath54 , which is sensitive to small halos at high redshifts , is more affected .
indeed , the contribution to @xmath54 by these halos is about @xmath149 , assuming a gas mass fraction of @xmath96 .
the finite box size and resolution also suppress power in the pressure power spectrum . as we saw in [ resolution ] and figure [ fig : pk_rhodm ] , the simulations lack power for @xmath150 and @xmath151 mpc@xmath35 . to test the impact of this suppression , we consider the constant bias model described in [ bias_model ] . the total sz power spectrum for this model
is shown in figure [ fig : cl_kran ] as the solid line , for the @xmath82cdm case .
this figure also shows the results of performing the same calculation , but after suppressing power in several ranges of @xmath109 values . the finite box size ( keeping only modes with @xmath152 mpc@xmath35 ) reduces @xmath153 slightly for @xmath154 and @xmath155 , and thus does not have a very large effect . on the other hand ,
the finite resolution ( @xmath156 mpc@xmath35 ) reduces @xmath153 considerably for @xmath157 .
the above results can be interpreted as follows . at a given @xmath41
, the limited @xmath109-range corresponds to a limited @xmath101-range , @xmath158 .
let us take @xmath159 and @xmath160 , as relevant for the simulations . then , @xmath161 , @xmath136 and @xmath162 correspond to @xmath163 , @xmath164 and @xmath165 , respectively . since most of contributions to @xmath166
come from @xmath167 , the finite @xmath168 decreases @xmath166 only at low @xmath41 , while @xmath169 does so over the entire @xmath41 range .
we conclude that the limitations of the simulations precludes us from predicting the sz power spectrum outside of the @xmath170 range .
these limits can only be improved by using larger simulations .
it is however worth noting that there could be more sz power around @xmath171 .
this might then be detectable by future interferometric cmb measurements that have angular resolutions around @xmath172 , intermediate in scale between the satellite missions and the planned millimeter experiments ( alma , lmsa ) .
the impact of secondary anisotropies on the upcoming map mission @xcite were studied by refregier et al.@xcite .
they showed that discrete sources , gravitational lensing and the sz effect were the dominant extragalactic foregrounds for map .
the dotted line on figure [ fig : cl ] shows the expected noise for measuring the primary cmb power spectrum with the 94 ghz map channel , with a band average of @xmath173 . for all model considered ,
the sz power spectrum is well below the noise .
the @xmath174 @xmath0-parameter for the map 94 ghz beam ( @xmath175 fwhm ) is listed in table [ tab : results ] for each model , from both the simulations and the ps formalism .
the resulting rms rj temperature fluctuations are of the order of a few @xmath176k , compared to a nominal antenna noise of about @xmath177k .
the sz effect will therefore not be a major limitation for estimating cosmological parameters with map . for comparison ,
the residual spectrum from undetected point sources ( @xmath178 jy ) expected using the model of toffolatti et al.@xcite is shown in figure [ fig : cl ] for the 94 ghz channel .
point sources dominate over the sz effect at @xmath179 , but are comparable below that .
moreover , we have shown in figure [ fig : cl_z ] and [ fig : cl_t ] , that about 50% of the sz power spectrum at @xmath180 is produced at low redshifts ( @xmath142 ) and by clusters of galaxies ( @xmath181 ) .
this confirms the results of refregier et al . , who predicted that most of the sz effect could be removed by cross - correlating the cmb maps with existing x - ray cluster catalogs ( eg .
xbacs @xcite , bcs @xcite ) .
because of its limited spectral coverage , the map mission will not permit a separation between the sz effect and primordial anisotropies .
apart from a handful of clusters which will appear as point sources , it will therefore be difficult to detect the sz fluctuations directly with map .
on the other hand , the future planck surveyor mission @xcite will cover both the positive and the negative side of the sz frequency spectrum , and will thus allow a clear separation of the different foreground and background components @xcite .
aghanim et al .
@xcite have established that , using such a separation , the sz profiles of individual clusters can be measured down to @xmath182 . moreover , hobson et al .
@xcite have estimated that the sz power spectrum could be measured for @xmath183 , with a precision per multipole of about 70% .
planck surveyor will therefore provide a precise measurement of the total sz power spectrum .
this would provide a direct , independent measurement of @xmath50 and of @xmath81@xcite , and would thus help breaking the degeneracies in the cosmological parameters estimated from primordial anisotropies alone . note that this measurement might be also feasible , albeit with less precision , with upcoming balloon experiments which also have broad spectral coverage .
[ planck ] the measured abundance of deuterium in low metallicity systems , together with big bang nucleosynthesis , predicts about twice as many baryons than what is observed in galaxies , stars , clusters and neutral gas @xcite .
these `` missing baryons '' are likely to be in the form of the warm gas in groups in filaments @xcite .
this component is indeed difficult to observe directly since it is too cold to be seen in the x - ray band , and too hot to produce any absorption lines in the quasar spectra @xcite .
the sz effect on large scale could however provides a unique probe of this warm gas .
one can indeed imagine subtracting the detected clusters from sz maps , and measuring the power spectrum of the residual sz fluctuations , which are mainly produced by groups and filaments .
for instance , if all clusters with @xmath184 kev were removed from the sz map , the sz power spectrum would drop by a factor of about 2 for @xmath134 ( see figure [ fig : cl_t ] ) . for the planck surveyor sensitivity quoted in [ planck ] , this yields a signal - to - noise ration per multipole of about 1 . the amplitude ,
if not the shape , of the residual sz spectrum will thus be easily detected by planck , thus yielding constraints on the temperature and density of the missing baryons . in our simulations , we have only included gravitational forces .
however , feedback from star and agn formation can also significantly heat the igm and thus affect the observed sz effect .
valegeas & silk @xcite ( see also reference therein ) , have studied the energy injection produced by photo - ionization , supernovae , and agn . in their model ,
agn are the most efficient , and can heat the igm by as much as @xmath185 k by a redshift of a few .
this results in a mean @xmath0-parameter of about @xmath2 , which is comparable to our value derived from gravitational instability alone .
preheating by feedback can thus increase the amplitude of the sz effect by a factor of a few , and thus be easily detected by planck .
feedback can thus be directly measured as an excess in the @xmath0-parameter or in the sz power spectrum , over the prediction from gravitational instability alone .
energy injection has a large effect on the gas in groups and filaments , comparatively to that in clusters .
we may thus also detect the effects of feedback through the relationship between the x - ray temperature of groups ( or their galaxy velocity dispersion ) and their sz temperature .
these measurements would then constrain the physics of energy injection .
we have studied the sz effect using mmh simulations .
our results for the mean comptonization parameter is consistent with earlier work using the press - schechter formalism and hydrodynamical simulations .
it is found to be lower than the current observational limit by about one order of magnitude , for all considered cosmologies .
the sz power spectrum is found to be comparable to the primary cmb power spectrum at @xmath186 . for the scdm model , our sz power spectrum is approximately consistent with that derived by persi et al.@xcite , after rescaling for the differing values of @xmath81 .
we found that groups and filaments ( @xmath187 ) contribute about 50% of the sz power spectrum at @xmath5 . on these scales , about 50% of the sz power spectrum is produced at @xmath6 and can thus be removed using x - ray cluster catalogs .
the sz fluctuations are well below the instrumental noise expected for the upcoming map mission , and should therefore not be a limiting factor .
the sz power spectrum should however be accurately measured by the future planck mission .
such a measurement will yield an independent measurement of @xmath50 and @xmath81 , and thus complement the measurements of primary anisotropies@xcite .
we have compared our simulation results with predictions from the ps formalism .
the results from the two methods agree approximately , but differ in the details .
the discrepancy could be due to the finite resolution of the simulations , which limit the validity of our predictions outside the @xmath1 range .
we also find discrepancies with other numerical simulations .
these issues can only be settled with larger simulations , and by a detailed comparison of different hydrodynamical codes .
such an effort is required for our theoretical predictions to match the precision with which the sz power spectrum will be measured in the future .
a promising approach to measure the sz effect on large scales is to cross - correlate cmb maps with galaxy catalogs @xcite .
most of the sz fluctuations on map s angular scales ( @xmath188 ) are produced at low redshifts and are thus correlated with tracers of the local large scale structure .
preliminary estimates indicate that such a cross - correlation between map and the existing apm galaxy catalog would yield a significant detection .
of course , even larger signals are expected for the planck surveyor mission .
this would again provide a probe of the gas distributed not only in clusters , but also in the surrounding large - scale structure and therefore help solve the missing baryon problem .
moreover , energy injection from star and agn formation can produce an sz amplitude in excess of our predictions , which only involve gravitational forces .
the measurement of sz fluctuations or of a cross - correlation signal thus provides a measure of feedback and can thus shed light on the process of galaxy formation .
we thank uros seljak and juan burwell for useful collaboration and exchanges .
we also thank renyue cen , greg bryan , jerry ostriker , arielle phillips and roman juszkiewicz for useful discussions and comparisons . a.r .
was supported in princeton by the nasa map / midex program and the nasa atp grant nag5 - 7154 , and in cambridge by an eec tmr grant and a wolfson college fellowship .
is partially supported by the map / midex program .
e.k . acknowledges a fellowship from the japan society for the promotion of science .
computing support from the national center for supercomputing applications is acknowledged .
u.p . was supported in part by nserc grant 72013704 .
j. e. carlstrom , m. k. joy , l. grego , g. p. holder , w. l. holzapfel , j. j. mohr , s. patel and e. d. reese , in nobel symposium particle physics and the universe , to appear in physica scripta and world scientific , eds .
l. bergstrom , p. carlson and c. fransson , preprint astro - ph/9905255 . | the hot gas in the igm produces anisotropies in the cosmic microwave background ( cmb ) through the thermal sunyaev - zeldovich ( sz ) effect .
the sz effect is a powerful probe of large - scale structure in the universe , and must be carefully subtracted from measurements of the primary cmb anisotropies .
we use moving - mesh hydrodynamical simulations to study the 3-dimensional statistics of the gas , and compute the mean comptonization parameter @xmath0 and the angular power spectrum of the sz fluctuations , for different cosmologies .
we compare these results with predictions using the press - schechter formalism .
we find that the two methods agree approximately , but differ in details .
we discuss this discrepancy , and show that resolution limits the reliability of our results to the @xmath1 range . for cluster normalized cdm models , we find a mean @xmath0-parameter of the order of @xmath2 , one order of magnitude below the current observational limits from the cobe / firas instrument . for these models ,
the sz power spectrum is comparable to the primordial power spectrum around @xmath3 .
it is well below the projected noise for the upcoming map satellite , and should thus not be a limitation for this mission .
it should be easily detectable with the future planck surveyor mission .
we show that groups and filaments ( @xmath4 kev ) contribute about 50% of the sz power spectrum at @xmath5 .
about half of the sz power spectrum on these scales are produced at redshifts @xmath6 , and can thus be detected and removed using existing catalogs of galaxies and x - ray clusters . 2 |
all graphs in this paper are finite , without loops and multiple edges . for a graph @xmath1
we denote by @xmath10 and @xmath11 the vertex set and the edge set of @xmath1 , respectively . by @xmath12
we denote the neighborhood of a vertex @xmath13 , i.e. the subset of vertices of @xmath1 adjacent to @xmath14 .
the subgraph of @xmath1 induced by a set @xmath15 will be denoted @xmath16 $ ] .
we say that a graph @xmath1 is @xmath17-free if @xmath1 does not contain @xmath17 as an induced subgraph . as usual ,
@xmath18 and @xmath19 stand for the complete graph and chordless path on @xmath20 vertices , respectively , and the complement of a graph @xmath1 is denoted @xmath21 .
a class of graphs , or synonymously graph property , @xmath22 is said to be _ hereditary _ if @xmath23 implies @xmath24 for any vertex @xmath13 .
we call @xmath22 _ monotone _ if @xmath23 implies @xmath24 for any vertex @xmath13 and @xmath25 for any edge @xmath26 .
clearly every monotone property is hereditary , but the converse statement is not true in general . a property @xmath22 is _ additive _ if @xmath27 and @xmath28 with @xmath29 implies @xmath30 .
the class of graphs containing no induced subgraphs isomorphic to graphs in a set @xmath31 will be denoted @xmath32 .
it is well known that a class of graphs @xmath22 is hereditary if and only if @xmath33 for some set @xmath31 .
a property is said to be non - trivial if it contains at least one , but not all graphs .
complementary property _ of @xmath22 is @xmath34 .
note that @xmath22 is hereditary if and only if @xmath35 is .
so a _ co - additive hereditary _ property , i.e. the complement of an additive hereditary property , is itself hereditary .
let @xmath0 be graph properties ( classes ) with @xmath36 .
a graph @xmath37 is _
@xmath38 @xmath39 @xmath40-colorable _ if there is a partition @xmath41 of @xmath10 such that @xmath42\in { { \cal p}}_j$ ] for each @xmath43 . the problem of recognizing @xmath38 @xmath39 @xmath40-colorable graphs is usually referred to as generalized graph coloring @xcite . when @xmath6 is the class @xmath44 of edgeless graphs , this problem coincides with the standard @xmath7-colorability , which is known to be np - complete for @xmath8 .
generalized graph coloring remains difficult for many other cases .
for example , cai and corneil @xcite showed that ( @xmath45,@xmath46)-coloring is np - complete for any integers @xmath47 , with the exception @xmath48 .
this result , and others @xcite , have been recently generalized in @xcite as follows .
[ thm : add ] if @xmath0 ( @xmath36 ) are additive hereditary classes of graphs , then the problem of recognizing @xmath38 @xmath39 @xmath40-colorable graphs is np - hard , unless @xmath49 and @xmath50 is the class of edgeless graphs
. clearly , a similar result follows for co - additive properties . in the present paper we focus on the case where we have a mixture of additive and co - additive properties .
the _ product _ of graph classes @xmath51 is @xmath52 is @xmath53-colorable@xmath54 .
a property is _ reducible _ if it is the product of two other properties , otherwise it is _ irreducible_. it can be easily checked that the product of additive hereditary ( or monotone ) properties is again additive hereditary ( respectively , monotone ) ; and that @xmath55 .
so , without loss of generality we shall restrict our study to the case @xmath49 and shall denote throughout the paper an additive property by @xmath22 and co - additive by @xmath56 .
we will refer to the problem of recognizing @xmath57-colorable graphs as @xmath58-recognition .
the plan of the paper is as follows . in section [ sec : hard ] , we show that @xmath58-recognition can not be simpler than @xmath22- or @xmath56-recognition . in particular
, we prove that @xmath58-recognition is np - hard whenever @xmath22- or @xmath56-recognition is np - hard .
then , in section [ sec : pol ] , we study the problem under the assumption that both @xmath22- and @xmath56-recognition are polynomial - time solvable and present infinitely many classes of @xmath57-colorable graphs with polynomial recognition time .
these two results together give a complete answer to the question of complexity of @xmath58-recognition when @xmath22 and @xmath59 are additive monotone .
when @xmath22 and @xmath59 are additive hereditary ( but not both monotone ) , there remains an unexplored gap that we discuss in the concluding section of the paper .
in this section we prove that if @xmath22-recognition ( or @xmath56-recognition ) is np - hard , then so is @xmath58-recognition .
this is a direct consequence of the theorem below . in this theorem
we use uniquely colorable graphs , which are often a crucial tool in proving coloring results . a graph @xmath1 is _
uniquely @xmath53-colorable _ if @xmath60 is its only @xmath53-partition , up to some permutation of the @xmath61 s .
if , say , @xmath62 , then @xmath63,@xmath64,@xmath65 will also be a @xmath66-coloring of @xmath1 ; such a permutation ( of @xmath61 s that correspond to equal properties ) is a _
trivial interchange_. a graph is _ strongly uniquely @xmath53-colorable _ if @xmath60 is the only @xmath53-coloring , up to trivial interchanges . when @xmath51 are irreducible hereditary properties , and each @xmath9 is either additive or co - additive , there is a strongly uniquely @xmath53-colorable graph with each @xmath61 non - empty .
this important construction , for additive @xmath9 s , is due to mihk @xcite , with some embellishments by broere and bucko @xcite , while the proof of unique colorability follows from ( * ? ? ?
* thm . 5.3 ) .
obviously , similar results apply to co - additive properties .
the generalization to mixtures of additive and co - additive properties can be found in ( * ? ? ?
* cor . 4.3.6 , thm .
5.3.2 ) [ thm : hard ] let @xmath22 and @xmath59 be additive hereditary properties . then there is a polynomial - time reduction from @xmath22-recognition to @xmath58-recognition .
let @xmath67 and @xmath68 , where the @xmath9 s and @xmath69 s are irreducible additive hereditary properties . as noted above
, there is a strongly uniquely @xmath70-colorable graph @xmath17 with partition @xmath71 @xmath39 @xmath72 , where each @xmath73 and @xmath74 is non - empty .
define @xmath75 and @xmath76 .
arbitrarily fix a vertex @xmath77 , and define @xmath78 . for any graph @xmath1 ,
let the graph @xmath79 consist of disjoint copies of @xmath1 and @xmath17 , together with edges @xmath80 .
we claim that @xmath81 if and only if @xmath82 .
if @xmath82 , then , by additivity , @xmath83 $ ] is in @xmath22 , and thus @xmath79 is in @xmath84 .
conversely , suppose @xmath81 , i.e. it has a @xmath85-partition , say @xmath86 . since @xmath17 is strongly uniquely partitionable , we can assume that , for @xmath87 , @xmath88 . now
, suppose for contradiction that , for some @xmath7 , there is a vertex @xmath89 such that @xmath90 ; without loss of generality , let @xmath91 .
then @xmath92 \cong h[w_r \cup \{u\}]$ ] is in @xmath93 , so @xmath94 is a new @xmath70-partition of @xmath17 , which is impossible .
thus , @xmath95 , and hence @xmath82 , as claimed . since @xmath17 is a fixed graph , @xmath79 can be constructed in time linear in @xmath96 , so the theorem is proved .
[ lem : pol ] for any @xmath97 and @xmath98 , there exists a constant @xmath99 such that for every graph @xmath100 and every subset @xmath101 with @xmath102\in { { \cal p}}$ ] , at least one of the following statements holds : * there is a subset @xmath103 such that @xmath104\in { { \cal p}}$ ] , @xmath105\in { { \cal q}}$ ] , and @xmath106 , * there is a subset @xmath107 such that @xmath108\in { { \cal p}}$ ] , @xmath109 , and @xmath110 . by the ramsey theorem @xcite , for each positive integers @xmath111 and @xmath20 , there is a constant @xmath112 such that every graph with more than @xmath112 vertices contains either a @xmath113 or a @xmath18 as an induced subgraph . for two classes @xmath97 and @xmath98 , we define @xmath99 to be equal @xmath112 .
let us show that with this definition the proposition follows .
let @xmath37 be a graph in @xmath114 , and @xmath115 a subset of @xmath116 such that @xmath102\in { { \cal p}}$ ] .
consider an arbitrary subset @xmath103 such that @xmath104\in { { \cal p}}$ ] and @xmath105\in { { \cal q}}$ ] . if ( a ) does not hold , then @xmath117 .
furthermore , @xmath118\in { { \cal p}}\cap { { \cal q}}\subseteq free(k_n,\overline{k}_m)$ ] , and hence @xmath119 . therefore , @xmath120 .
but then any subset @xmath121 such that @xmath122 and @xmath109 satisfies ( b ) .
lemma [ lem : pol ] suggests the following recognition algorithm for graphs in the class @xmath114 .
a graph @xmath37 . * output : * * yes *
if @xmath123 , or * no * otherwise . * find in @xmath1 any inclusion - wise maximal subset @xmath101 inducing a @xmath18-free graph .
* if there is a subset @xmath107 satisfying condition ( b ) of lemma [ lem : pol ] , + then set @xmath124 and repeat step ( 2 ) .
* if @xmath1 contains a subset @xmath103 such that * * @xmath119 , * * @xmath106 , * * @xmath104\in { { \cal p}}$ ] , * * @xmath105\in { { \cal q}}$ ] , + output * yes * , otherwise output * no*. [ thm : pol ] if graphs on @xmath125 vertices in a class @xmath97 can be recognized in time @xmath126 and graphs in a class @xmath98 can be recognized in time @xmath127 , then algorithm @xmath128 recognizes graphs on @xmath125 vertices in the class @xmath84 in time @xmath129 , where @xmath99 .
correctness of the algorithm follows from lemma [ lem : pol ] .
now let us estimate its time complexity . in step ( 2 )
, the algorithm examines at most @xmath130@xmath131 subsets @xmath132 and for each of them verifies whether @xmath108\in { { \cal p}}$ ] in time @xmath126 .
since step ( 2 ) loops at most @xmath125 times , its time complexity is @xmath133 . in step ( 3 )
, the algorithm examines at most @xmath130@xmath134 subsets @xmath135 , and for each @xmath135 , it verifies whether @xmath104\in { { \cal p}}$ ] in time @xmath126 and whether @xmath105\in { { \cal q}}$ ] in time @xmath136 .
summarizing , we conclude that the total time complexity of the algorithm is @xmath129 .
notice that theorem [ thm : pol ] generalizes several positive results on the topic under consideration .
for instance , the split graphs @xcite , which are ( @xmath137,@xmath138)-colorable by definition , can be recognized in polynomial time .
more general classes have been studied under the name of polar graphs in @xcite . by definition ,
a graph is @xmath139 polar if it is @xmath140-colorable with @xmath141 and @xmath142 .
it is shown in @xcite that for any particular values of @xmath143 and @xmath144 , @xmath139 polar graphs on @xmath125 vertices can be recognized in time @xmath145 .
further examples generalizing the split graphs were examined in @xcite and @xcite , where the authors showed that classes of graphs partitionable into at most two independent sets and two cliques can be recognized in polynomial time . these are special cases of @xmath58-recognition with @xmath146 and @xmath147 .
theorems [ thm : hard ] and [ thm : pol ] together provide complete answer to the question of complexity of @xmath58-recognition in case of monotone properties @xmath22 and @xmath59 .
indeed , if @xmath22 is an additive monotone non - trivial property , then @xmath97 for a certain value of @xmath20 , since otherwise it includes all graphs . similarly , if @xmath59 is additive monotone , then @xmath98 for some @xmath111 . hence , the following theorem holds .
[ co - her - additive ] if @xmath22 and @xmath59 are additive monotone properties , then @xmath148-recognition has polynomial - time complexity if and only if @xmath22- and @xmath56-recognition are both polynomial - time solvable ; moreover , @xmath148-recognition is in np if and only if @xmath22- and @xmath56-recognition are both in np .
if @xmath22 and @xmath59 are general additive hereditary properties ( not necessarily monotone ) , then there is an unexplored gap containing properties @xmath84 , where @xmath22 and @xmath56 can both be recognized in polynomial time , but @xmath149 or @xmath150 ( where @xmath151 is the set of cliques ) . in the rest of this section
we show that this gap contains both np - hard and polynomial - time solvable instances , and propose several open problems to study .
for a polynomial time result we refer the reader to @xcite , where the authors claim that @xmath58-recognition is polynomial - time solvable if @xmath22 is the class of edgeless graphs and @xmath152 .
notice that @xmath153 contains all edgeless graphs and hence theorem [ thm : pol ] does not apply to this case .
interestingly enough , when we extend @xmath22 to the class of bipartite graphs , we obtain an np - hard instance of the problem , as the following theorem shows .
[ thm : ex1 ] if @xmath22 is the class of bipartite graphs and @xmath152 , then @xmath58-recognition is np - hard .
we reduce the standard 3-colorability to our problem .
consider an arbitrary graph @xmath1 and let @xmath154 be the graph obtained from @xmath1 by adding a triangle @xmath155 with no edges between @xmath1 and @xmath156 .
we claim that @xmath1 is 3-colorable if and only if @xmath154 is @xmath157-colorable .
first , assume that @xmath1 is 3-colorable and let @xmath158 be a partition of @xmath10 into three independent sets .
we define @xmath159 for @xmath160 .
then @xmath161 $ ] is a bipartite graph and @xmath162\in free(\overline{p}_3)$ ] , and the proposition follows .
conversely , let @xmath163 be a partition of @xmath164 with @xmath165 $ ] being a bipartite graph and @xmath166\in free(\overline{p}_3)$ ] .
clearly , @xmath167 .
if @xmath168 contains a single vertex , then @xmath169 $ ] is an edgeless graph , since otherwise a @xmath170 arises . if @xmath168 contains more than one vertex , then @xmath171 for the same reason .
clearly , in both cases @xmath1 is a 3-colorable graph .
this discussion presents the natural question of exploring the boundary that separates polynomial from non - polynomial time solvable instances in the above - mentioned gap .
as one of the smallest classes in this gap with unknown recognition time complexity , let us point out @xmath157-colorable graphs with @xmath172 and @xmath173 , where @xmath174 is the disjoint union of two copies of @xmath175 .
another direction for prospective research deals with @xmath140-colorable graphs where @xmath22 or @xmath56 is neither additive nor co - additive .
this area seems to be almost unexplored and also contains both np - hard and polynomial - time solvable problems . to provide some examples ,
let @xmath56 be the class of complete bipartite graphs , which is obviously neither additive nor co - additive .
the class of graphs partitionable into an independent set and a complete bipartite graph has been studied in @xcite under the name of bisplit graphs and has been shown there to be polynomial - time recognizable .
again , extension of @xmath22 to the class of all bipartite graphs transforms the problem into an np - hard instance .
[ thm : ex2 ] if @xmath22 is the class of bipartite graphs and @xmath56 is the class of complete bipartite graphs , then @xmath58-recognition is np - hard .
the reduction is again from 3-colorability . for a graph @xmath1
, we define @xmath154 to be the graph obtained from @xmath1 by adding a new vertex adjacent to every vertex of @xmath1 .
it is a trivial exercise to verify that @xmath1 is 3-colorable if and only if @xmath154 is @xmath157-colorable .
the second author wishes to thank r. bruce richter , his doctoral supervisor , for his valuable comments ; as well as the canadian government , which is fully funding his studies in waterloo through a canadian commonwealth scholarship . ,
uniqueness and complexity in generalised colouring .
thesis , university of waterloo , waterloo , ontario , canada .
february 2003 .
( available on - line at http://www.math.uwaterloo.ca/@xmath179afarrugia/thesis.ps or http://etheses.uwaterloo.ca ) | for graph classes @xmath0 , generalized graph coloring is the problem of deciding whether the vertex set of a given graph @xmath1 can be partitioned into subsets @xmath2 so that @xmath3 induces a graph in the class @xmath4 @xmath5 .
if @xmath6 is the class of edgeless graphs , then this problem coincides with the standard vertex @xmath7-colorability , which is known to be np - complete for any @xmath8 .
recently , this result has been generalized by showing that if all @xmath9 s are additive induced - hereditary , then the generalized graph coloring is np - hard , with the only exception of bipartite graphs .
clearly , a similar result follows when all the @xmath9 s are co - additive . in this paper
, we study the problem where we have a mixture of additive and co - additive classes , presenting several new results dealing both with np - hard and polynomial - time solvable instances of the problem .
_ keywords : _ generalized graph coloring ; polynomial algorithm ; np - completeness |
the three - body and four - body decays of @xmath8 meson are more complicated than two - body case , specifically where both non - resonant and resonant contribution exist . in this
types the theoretical momentums of the output particles are not directly calculable . in four - body cases ,
the momentums and form factors are written in terms of five variables : two invariant - mass of particles variable and three angles .
+ in the previous works @xcite , for some specific three - body decays assumed that one of the output mesons due to its heavy weigh , remains stationary and the two other output mesons move back to back and the amplitude of this three decay modes can be calculate . in this work we extended this assumption to
non - resonant four - body decays and calculate the branching ratio of @xmath0 decay in naive factorization approach where we can ignore the non - factorization effects .
we assume that the @xmath5 output meson , due to its heavy weight , remains stationary and the three pion mesons move in a plane and we have directly resolved their momentums and we calculated the branching ratio of @xmath9 and predicted @xmath6 which is compatible with the experimental result @xmath10 .
+ this paper is organized as follows . in section 2
we consider the amplitudes of the @xmath9 decay and we present the transition matrix needed and the kinematics of the four - body decays . in section 3
we give the numerical results , and in the last section we have a summary .
for the four - body @xmath11 , the feynman diagrams are shown in fig .
1 , where both panel show the @xmath12 tree transitions @xcite . under the factorization approach ,
the contribution of the diagram ( a ) in decay amplitude consists of the @xmath13 distinct factorization term , where @xmath14 denotes the form factor of @xmath8 to @xmath5 meson transitional , and the contribution of the diagram ( b ) consists of a @xmath15 distinct factorization term , where @xmath16 denotes the two - meson transition matrix element . decay . ]
so the matrix elements of the @xmath11 decay amplitudes are given by @xmath17 where in factorization , the general expression of the @xmath18 quantities in terms of effective wilson coefficients is @xcite @xmath19 according to fig .
1(a ) , the form factor of @xmath8 meson to @xmath5 meson is defined as fallow @xcite @xmath20}f_1(q^2 ) \nonumber \\ & & + \frac{m_b^2-m_d^2}{(p_b - p_4)^2}(p_b - p_4)_\mu f_0(q^2),\end{aligned}\ ] ] and the point like 3-body matrix element is assumed to be @xcite @xmath21}. \nonumber \\\end{aligned}\ ] ] according to fig .
1(b ) , for the current - induced process , the two - meson transition matrix element @xmath22 has the fallowing general expression @xcite @xmath23 where the form factors @xmath24 and @xmath25 for the non - resonant decay are given by @xmath26 @xmath27}+\frac{f_b}{2f_\pi^2},\end{aligned}\ ] ] @xmath28}.\end{aligned}\ ] ] the tow - body @xmath29 matrix element can be related to the @xmath5 and @xmath30 matrix element of the weak interaction current : @xmath31 where @xmath32 is the @xmath5 to @xmath30 transition form factor and determined from experiment @xcite . for four - body @xmath8
decays , we can work in the rest frame of the b meson . by following the definition given in @xcite ,
when b meson decays to final - state mesons , the kinematics of the four - body decays can be described in term of five independent variables where the momenta and masses are defined as @xmath33 with @xmath34 and @xmath35 where @xmath36 and @xmath37 .
we choose the set of independent variables as @xmath38 which have the following geometrical meaning ( see fig .
( 2 ) ) * @xmath39 , is the invariant - mass of particles 1 and 2 ; * @xmath40 , is the invariant - mass of particles 3 and 4 ; * @xmath41 , is the angle between the three - momentum of particle 1 ( particle 3 ) with respect to the direction of @xmath42 ( @xmath43 defined in the rest frame of the decaying particle ( see fig.(2 ) ) ; * @xmath44 is the angle between the planes defined by particles ( 1 , 2 ) and ( 3 , 4 ) also in the rest frame of the decaying particle . .
we have defined @xmath45 , such that @xmath46 . ] with this choice of kinematics , the differential decay rate in the rest frame of the decaying particle can be written as : @xmath47 where @xmath48 ) is the velocity of particle 1 ( particle 3 ) in the center of mass frame of particles 1 and 2 ( 3 and 4 ) and @xmath49^{1/2}$ ] , the @xmath50 factor accounts for identical particles in the phase space and @xmath51 is the amplitude of the decay ( eq.(1 ) ) . + for considering the @xmath0 decay mode , we work in the rest frame of the @xmath8 meson and assume that the @xmath52 meson due to its heavy weight , has a small momentum and can easily be neglected , so the three pions move in a plane . hence the momentums of @xmath8 meson and final - state mesons are given by @xmath53 where @xmath54 and @xmath55 .
now we can easily integrate the decay rate with respect to all independent variables and we obtain a simple experiment @xmath56
the wilson coefficients @xmath57 have been calculated in different schemes . in this paper
we use consistently the naive dimensional regularization ( ndr ) scheme .
the values of @xmath57 at the scales @xmath58 at the leading order ( lo ) are @xcite @xmath59 the parameter @xmath60 in the form factors , determined from the @xmath61 decay is @xmath62 for the elements of the ckm matrix , we use the values of the wolfenstein parameters and obtain @xcite @xmath63 the meson masses , decay constants and form factors needed in our calculations are taken as ( in units of gev ) @xcite @xmath64 using the parameters relevant for the @xmath0 decay , we calculate the branching ratio of this decay which is @xmath65 where the experimental value is @xcite @xmath66
in this work , we have considered four - body @xmath67 decays under the assumption that the @xmath5 meson has small and negligible momentum due to its heavy weight .
this decay mode have contribution of @xmath68 and @xmath69 coefficients . according to the feynman diagrams ,
each contribution factorized into a form factor and a decay constant . under the assumption of negligible momentum of d meson ,
the three output pion mesons move in a plane and we have directly resolved their momentums and we calculated the branching ratio of @xmath9 and predicted @xmath6 which is compatible with the experimental result @xmath10 .
10 behnam mohammadi and hossein mehraban,_``study of the @xmath70 decay '' _ , phys.rev .
d * 89*(9):095026 , 2014 . behnam mohammadi and hossein mehraban .
_ `` three - body decays of @xmath71''_. eur.phys.j .
a**50**:122 , 2014 .
michael gronau , oscar f. hernandez , david london , and jonathan l. rosner .
_ `` broken su(3 ) symmetry in two - body b decays''_. phys.rev .
d**52**:63566373 , 1995 .
mahiko suzuki .
_ `` selection rules in three - body b decay from factoriza - tion''_. phys.rev .
d**62**:091502 , 2000 .
s. fajfer , r.j .
oakes , and t.n .
_ `` the penguin operators in non - resonant @xmath72 decays''_. phys.rev .
d**60**:054029 , 1999 .
k. terasaki .
_ `` nonfactorizable long distance contributions in color sup - pressed decays of b mesons''_. int.j.mod.phys .
a * 13*:43254338 , 1998 .
a. ali , c. greub , _ `` an analysis of two - body non - leptonic b decays involving light mesons in the standard model''_. phys.rev.d . * 57*:2996 - 3016 , 1998 .
[ arxiv : hep - ph/9707251 ] hai - yang cheng and kwei - chou yang .
_ `` nonresonant three - body decays of d and b mesons''_. phys.rev .
d * 66*:054015 , 2002 . b. bajc , s. fajfer , r.j .
oakes , t.n .
pham , and sasa prelovsek .
_ `` the cp violating asymmetry in @xmath73 decays''_. phys.lett .
b * 447*:313 320 , 1999 .
tomoko iwashita .
_ `` search for a @xmath74 resonance in@xmath75 decay at belle''_. 2012 .
n. quintero , g. lopez castro , and d. delepine .
_ `` lepton number violation in top quark and neutral b meson decays''_. phys.rev .
d * 84*:096011,2011 .
gabriel lopez castro and nestor quintero .
_ `` lepton number violating four - body tau lepton decays''_. phys.rev .
d * 85*:076006 , 2012 .
m. beneke , g. buchalla , m. neubert , and christopher t. sachrajda .
_ `` qcd factorization in @xmath76 decays and extraction of wolfenstein parameters''_. nucl.phys .
b * 606*:245321 , 2001 .
claude amsler et al .
review of particle physics .
b * 667*:11340 , 2008 . k.a .
olive et al .
review of particle physics .
c * 38*:090001 , 2014 .
ying - ying fan , wen - fei wang , shan cheng , and zhen - jun xiao .
_ `` semileptonic decays @xmath77 in the perturbative qcd factorization approach''_. chin.sci.bull . * 59*:125132 , 2014 . | we calculate the branching ratio of the non - resonant @xmath0 decay using a simple model based on the framework of the factorization approach where one can ommit the nonfactorization effects . in naive factorization approach
, there are only two tree diagrams for this decay mode . in the first diagram ,
the matrix element of decay mode is factorized into a @xmath1 form factor multiplied by a @xmath2 decay constant and in the second diagram , the matrix element is factorized into a @xmath3 form factor multiplied by a @xmath4 decay constant , we assume that in the rest frame of b meson , the @xmath5 meson remains stationary , so we obtain the value @xmath6 for the branching ratio of the @xmath0 decay mode , while the experimental results are @xmath7 . |
the recent experimental situation in hadron physics enables us to utilize huge amounts of data on heavy hadrons , which contain charm or bottom quark(s ) , for the investigation of hadron structures .
especially , decay properties of heavy mesons can shed more light on the nature of the light scalar mesons [ @xmath3 , @xmath16 , @xmath5 , and @xmath6 , which has been a hot topic in hadron physics @xcite .
for instance , the decay @xmath17 has been experimentally measured in refs .
@xcite for the study of the @xmath3 and @xmath5 resonances , and they observed a pronounced peak for the @xmath5 while no evident signal was found for the @xmath3 .
then a theoretical study @xcite followed the experiments and reproduced ratios of experimental branching fractions at a quantitative level , pointing out that @xmath18 production in the @xmath19 decay and a hadronization of @xmath20 to @xmath21 are essential to understand the branching fractions of the @xmath19 decay into @xmath22 . in the theoretical study
, the final state interaction between two pseudoscalar mesons is calculated with the so - called chiral unitary approach @xcite , in which the light scalar mesons are obtained as dynamically generated resonances , and it is concluded that the @xmath5 has a substantial fraction of the strange quarks . the same hadronization scheme has been employed in theoretical studies in refs .
@xcite . into @xmath23 and a primary @xmath20 pair .
( b ) semileptonic decay of @xmath0 into @xmath23 and two pseudoscalar mesons @xmath24 with a hadronization.,title="fig : " ] in this paper , we consider the semileptonic decay of @xmath25 , extending a discussion for the semileptonic @xmath26 decays into @xmath27 and @xmath28 resonances in ref .
the semileptonic @xmath11 decays have been experimentally investigated in , _
e.g. _ , bes @xcite , focus @xcite , babar @xcite , and cleo @xcite . here , in order to grasp how the semileptonic decay takes place , let us consider the @xmath0 meson . since the constituent quark component of @xmath0 is @xmath29 , we expect a cabibbo favored semileptonic decay of @xmath30 and hence the decay @xmath31 with @xmath20 being the vector meson @xmath32 , which is depicted in fig . [ fig:12](a )
actually this semileptonic decay mode has been observed in experiments , and its branching fraction to the total decay width is @xmath33 = 2.49 \pm 0.14 \%$ ] @xcite ( see table [ tab : br ] , in which we list branching fractions for the semileptonic decays of @xmath0 , @xmath1 , and @xmath2 reported by the particle data group ) . on the other hand , we can not straightforwardly extend the discussion to the scalar meson productions in the final state of the semileptonic decays , since the structure of the scalar mesons , whether @xmath34 or some exotic one , is still controversial . in this study
we consider the production of the @xmath5 or @xmath3 as dynamically generated resonances in the semileptonic @xmath0 decay , so we have to introduce an extra @xmath35 pair to make a hadronization as shown in fig . [ fig:12](b ) .
the introduction of an extra @xmath35 pair to make a hadronization has been performed in refs .
@xcite . in this study
we apply the same method of the hadronization to the semileptonic decays of @xmath11 mesons so as to investigate the nature of the light scalar mesons .
8.6cm@lc + mean life [ s ] & @xmath36 + @xmath37 $ ] & @xmath38 + @xmath39 $ ] & @xmath40 + @xmath41 $ ] & @xmath42 + @xmath43 $ ] & @xmath44 + + + mean life [ s ] & @xmath45 + @xmath46 $ ] & @xmath47 + @xmath48 $ ] & @xmath49 + @xmath50 $ ] & @xmath51 + @xmath52 $ ] & @xmath53 + @xmath54 $ ] & @xmath55 + @xmath39 $ ] & @xmath56 + @xmath37 $ ] & @xmath57 + + + mean life [ s ] & @xmath58 + @xmath59 $ ] & @xmath60 + @xmath61 $ ] & @xmath62 + @xmath63 $ ] & @xmath64 + @xmath65 $ ] & @xmath66 + @xmath67 $ ] & @xmath68 + utilizing the semileptonic decay of a heavy hadron provides us with two advantages when we investigate the internal structure of hadrons in the final state of the semileptonic decay .
first , cabibbo favored and suppressed processes enable us to specify flavors of quarks contained in final state hadrons .
second , the semileptonic decay of the heavy hadron to two light hadrons @xmath69 brings a suitable condition to measure effects of the final state interaction of the two light hadrons , since the leptons and hadrons in the final state interact with each other only weakly .
theoretical work on the issues of the semileptonic @xmath11 decays is already available . in ref .
@xcite , using qcd sum rules , the @xmath0 and @xmath1 semileptonic decays into @xmath70 are considered concluding that the importance of up and down quarks in the @xmath5 is not negligible . in ref .
@xcite the @xmath71 reaction is analyzed from the point of view of the @xmath5 being a @xmath34 state , concluding that @xmath20 component of the @xmath5 may not be dominant . in ref .
@xcite the @xmath72 reaction is studied concluding that it supports the dominant four quark nature of the @xmath3 and @xmath5 .
similar conclusions about the four quark nature of the scalar mesons are reached in the work of @xcite .
research along the same line is done in ref .
@xcite , looking for likely reasonable ratios that would help distinguish between the two and four quark structure of the scalar mesons .
another line of research is done using light - cone sum rules to evaluate the form factors appearing in the process @xcite .
this line of research is applied in many related processes , rare decays like @xmath73 in @xcite , @xmath74 in @xcite , @xmath75 and @xmath76 decays in @xcite , or semileptonic decays @xcite . in some cases the meson final state interaction is further implemented using the omnes representation @xcite , while in other cases breit wigner or flatte structures are implemented and parametrized to account for the resonances observed in the experiment .
in contrast to these pictures , in the present study we treat the scalar mesons as dynamically generated resonances from two pseudoscalar mesons in the so - called chiral unitary approach . then we describe the semileptonic decays of @xmath11 mesons in an economical way for hadronization as done in refs .
@xcite .
this paper is organized as follows . in sec .
[ sec : form ] we formulate the semileptonic decay widths of @xmath0 , @xmath1 , and @xmath2 into the light scalar and vector mesons and give our model of the hadronization .
we also calculate meson meson scattering amplitudes to generate dynamically the scalar mesons . in sec .
[ sec : results ] we show our numerical results of the semileptonic decay widths of @xmath0 , @xmath1 , and @xmath2 .
we predict branching fractions which are not reported by the particle data group and show invariant mass distributions of the two pseudoscalar mesons from the scalar and vector mesons . section [ sec : conclusion ] is devoted to drawing the conclusion of this study .
8.6cm@lc + @xmath77 & favored + @xmath78 & suppressed + @xmath79 & favored + @xmath80 & favored + @xmath81 & suppressed + + + @xmath82 & favored + @xmath83 & suppressed + @xmath84 & suppressed + @xmath79 & suppressed + @xmath85 & suppressed + @xmath80 & suppressed + @xmath86 & favored + + + @xmath87 & favored + @xmath88 & suppressed + @xmath89 & suppressed + @xmath90 & suppressed + @xmath91 & favored + in this section we formulate the semileptonic decay widths of @xmath0 , @xmath1 , and @xmath2 into light scalar and vector mesons : @xmath92 where @xmath93 , @xmath94 , and @xmath24 represent the light scalar , vector , and pseudoscalar mesons , respectively , and the lepton flavor @xmath95 can be @xmath96 and @xmath97 .
explicit decay modes are listed in table [ tab : mode ] . in order to formulate the decay width , we consider first the semileptonic decay amplitudes and widths in section [ sec : decay ] and next hadronizations into scalar and vector mesons in section [ sec : had ] . scattering amplitudes of two pseudoscalar mesons are then constructed in the chiral unitary approach for the description of the scalar mesons in section [ sec : chua ] . throughout this study
we assume isospin symmetry for light hadrons . in general , we can express the decay amplitude of @xmath25 , @xmath99 , by using the propagator of the @xmath100 boson and its couplings to leptons and quarks , which can be replaced with the fermi coupling constant @xmath101 . at this stage
we do not fix the number of the final state hadrons . in a similar manner to the formulation in ref .
@xcite , the explicit form of @xmath99 becomes @xmath102 the factor @xmath103 consists of the wave function of quarks inside the @xmath11 meson , the hadronization contribution in the final state , and the cabibbo
maskawa matrix element for the transition from the charm to a light quark .
the explicit form of @xmath103 will be determined in the next subsection .
the lepton and quark parts of the @xmath100 boson couplings are defined as .
@xmath104 respectively , where @xmath105 , @xmath106 , @xmath107 , and @xmath108 are the dirac spinors corresponding to the neutrino , lepton @xmath109 , light quark @xmath110 , and charm quark , respectively .
let us now calculate the squared amplitude for the semileptonic @xmath11 decay widths , in which we average ( sum ) the polarizations of the initial - state quarks ( final state leptons and quarks ) .
therefore , in terms of the amplitude in eq . , we can obtain the squared decay amplitude as @xmath111 where the factor @xmath112 comes from the average of the charm quark polarization in the initial state .
we can further calculate the lepton and quark parts in the amplitude , by using the conventions of the dirac spinors and traces of dirac @xmath113 matrices summarized in appendix [ app:1 ] , which lead to @xmath114 \notag \\ = & 2 \frac{p_{l}^{\alpha } p_{\nu}^{\beta } + p_{\nu}^{\alpha } p_{l}^{\beta } - p_{l } \cdot p_{\nu } g^{\alpha \beta } + i \epsilon ^{\alpha \beta \rho \sigma } p_{l \rho } p_{\nu \sigma}}{m_{l } m_{\nu } } , \end{aligned}\ ] ] where @xmath115 and @xmath116 ( @xmath117 and @xmath118 ) are momenta ( masses ) of the lepton @xmath109 and neutrino , respectively , and @xmath119 \notag \\ = & 2 \frac{p_{c \alpha } p_{q \beta } + p_{q \alpha } p_{c \beta } - p_{c } \cdot p_{q } g_{\alpha \beta } + i \epsilon _ { \alpha \beta \rho \sigma } p_{c}^{\rho } p_{q}^{\sigma}}{m_{c } m_{q } } , \end{aligned}\ ] ] with the momenta ( masses ) of the charm and light quarks , @xmath120 and @xmath121 ( @xmath122 and @xmath123 ) , respectively . is for a quark in the primary @xmath34 pair after the @xmath100 boson emission , which means that the momentum @xmath121 is carried by the constituent quark .
accordingly , @xmath123 is the mass of the constituent quark rather than of the current quark . in this sense
, @xmath123 respects the flavor @xmath124 symmetry . ]
then with a straightforward calculation we have @xmath125 now let us rewrite the momenta of quarks by using those of hadrons in the following manner : @xmath126 where we have neglected the relative internal momenta of the quarks , which are typically small compared to the masses of quarks . here
@xmath127 and @xmath128 ( @xmath129 and @xmath130 ) are the masses ( momenta ) of the @xmath11 and @xmath131 , @xmath94 mesons , respectively . with these translations
the square of @xmath132 with polarization summation becomes @xmath133 therefore , we obtain the squared decay amplitude as : @xmath134 with the above squared amplitude we can compute the decay width .
we will be interested in two types of decays : three - body decays for vector mesons such as @xmath135 , and four - body decays for scalar mesons constructed from two pseudoscalar mesons such as @xmath136 .
as it will be seen , both decay types can be described by the amplitude @xmath99 with different assumptions for @xmath103 : @xmath137 and @xmath138 respectively .
the formula for the three - body decay is given by @xcite : @xmath139 where @xmath140 is the momentum of the final state vector meson in the @xmath11 rest frame and @xmath141 is the momentum of the neutrino in the @xmath142 rest frame , both of which are evaluated as @xmath143^{2 } , \ , m_{v}^{2 } ) } { 2 m_{d } } , \ ] ] @xmath144^{2 } , \
, m_{l}^{2 } , \ , m_{\nu}^{2 } ) } { 2 m_{\rm inv}^ { ( l \nu ) } } , \label{eq : pnu_tilde}\ ] ] with the function @xmath145 and the vector meson mass @xmath146 .
the tilde on characters for leptons indicates that they are evaluated in the @xmath142 rest frame .
the solid angles @xmath147 and @xmath148 are for the vector meson in the @xmath11 rest frame and for the neutrino in the @xmath142 rest frame , respectively , and @xmath149 is the @xmath142 invariant mass .
the integral range of @xmath150 is @xmath151 $ ] . substituting the squared amplitude with that in eq .
, we obtain @xmath152 in general , the hadronization part @xmath103 may depend on the energy and scattering angles , and hence one can not put it out of the integral . in this study , however , @xmath137 will be simply constructed , so that this will not depend on @xmath150 nor the angle , as we will see in the next subsection .
furthermore , the integral of the solid angle @xmath153 is performed in the @xmath142 rest frame as @xcite @xmath154^{2 } \left ( \tilde{e}_{d } \tilde{e}_{v } - \frac{1}{3 } | \tilde{\bm{p}}_{d } |^{2 } \right ) , \end{aligned}\ ] ] where @xmath155 and @xmath156 are the energies and momenta in the @xmath142 rest frame . at the first equality we have used relations @xmath157 and @xmath158 , while at the third equality we have used relations obtained by neglecting masses of leptons : latexmath:[\[\tilde{e}_{l } = \tilde{e}_{\nu } = energies and momentum of hadrons in the @xmath142 rest frame can be exactly evaluated as @xmath160^{2 } - m_{v}^{2}}{2 m_{\rm inv}^ { ( l \nu ) } } , \ ] ] @xmath161^{2 } - m_{v}^{2}}{2 m_{\rm inv}^ { ( l \nu ) } } , \ ] ] and @xmath162 . as a consequence
, we have @xmath163^{2 } \notag \\ & \times \left ( \tilde{e}_{d } \tilde{e}_{v } - \frac{1}{3 } | \tilde{\bm{p}}_{d } |^{2 } \right ) , \end{aligned}\ ] ] where we have performed the integral of the solid angle @xmath147 . in a similar way
, we can evaluate the decay width for the four - body final state .
the formula for the four - body decay is given by @xmath164 where @xmath165 is the invariant mass of the two - meson system ( @xmath166 ) , @xmath167 is the center - of - mass momentum of the two - meson system in the @xmath11 rest frame and @xmath168 is the momentum of a meson in the @xmath166 rest frame , both of which are evaluated as @xmath169^{2 } , \ , [ m_{\rm inv}^ { ( l \nu ) } ] ^{2})}{2 m_{d } } , \ ] ] @xmath170^{2 } , \ ,
m_{h}^{2 } , \ , m_{h}^{\prime 2 } ) } { 2 m_{\rm inv}^{(h h ) } } , \ ] ] with the meson masses @xmath171 and @xmath172 . the momentum of the neutrino in the @xmath142 rest frame @xmath141 is given in eq . .
the solid angles @xmath173 and @xmath174 are for the two - meson system in the @xmath11 rest frame and for a meson in the @xmath166 rest frame , respectively .
the tilde on characters for mesons indicates that they are evaluated in the @xmath166 rest frame .
since we are interested in the meson meson invariant mass distributions for the semileptonic @xmath11 decay , we calculate the differential decay width @xmath175 . then in a similar manner to the case of the three - body decay
, we have @xmath176^{2 } \notag \\ & \times \left ( \tilde{e}_{d } \tilde{e}_{s } - \frac{1}{3 } | \tilde{\bm{p}}_{d } |^{2 } \right ) , \end{aligned}\ ] ] where we have performed the integrals with respect to the solid angles @xmath173 and @xmath174 .
we mention that @xmath177 will be simply constructed as well , so that this can be put out of the integral , as we will see in the next subsection .
the two - meson invariant mass @xmath165 can take a value within @xmath178 $ ] , while the integral range of @xmath150 is @xmath179 $ ] .
the energies and momentum of hadrons in the parentheses can be exactly evaluated as @xmath160^{2 } - [ m_{\rm inv}^{(h h ) } ] ^{2}}{2 m_{\rm inv}^ { ( l \nu ) } } , \ ] ] @xmath180^{2 } - [ m_{\rm inv}^{(h h ) } ] ^{2}}{2 m_{\rm inv}^ { ( l \nu ) } } , \ ] ] and @xmath162 .
next we fix the mechanism for the appearance of the scalar and vector mesons in the final state of the semileptonic decay .
we here note that , for the scalar and vector mesons in the final state , the hadronization processes should be different from each other according to their structure . for the scalar mesons , we employ the chiral unitary approach @xcite , in which the scalar mesons are dynamically generated from the interaction of two pseudoscalar mesons governed by the chiral lagrangians .
therefore , in this picture the light quark
antiquark pair after the @xmath100 boson emission gets hadronized by adding an extra @xmath35 with the quantum number of the vacuum , @xmath181 , which results in two pseudoscalar mesons in the final state [ see fig .
[ fig:12](b ) ] .
then the scalar mesons are obtained as a consequence of the final state interaction of the two pseudoscalar mesons as diagrammatically shown in fig .
[ fig:3 ] . for the vector mesons , on the other hand , hadronization with an extra @xmath35 is unnecessary since they are expected to consist genuinely of a light quark
antiquark pair [ see fig .
[ fig:12](a ) ] .
first we consider processes with the scalar mesons in the final state as the dynamically generated resonances .
the basic idea of the hadronization with an extra @xmath35 with the quantum number of the vacuum has already shown in refs .
we start with the @xmath182 matrix @xmath183 : @xmath184 one can easily check that this matrix has the property @xmath185 with this property , a @xmath186 pair after the @xmath100 boson emission can be added by an extra @xmath35 to be @xmath187 where @xmath188 denotes the flavor of light quarks : @xmath189 , @xmath190 , and @xmath191 .
next we rewrite the matrix @xmath183 in terms of the matrix @xmath192 for pseudoscalar mesons @xmath193 where we have taken into account the @xmath194@xmath195 mixing in a standard way @xcite . in this scheme
we can calculate the weight of each pair of pseudoscalar mesons in the hadronization .
namely , the @xmath20 pair gets hadronized as @xmath196 , where @xmath197 here and in the following we omit the @xmath195 contribution since @xmath195 is irrelevant to the description of the scalar mesons due to its large mass . in similar manners ,
the @xmath198 , @xmath199 , @xmath200 , @xmath201 , and @xmath202 pairs get hadronized as @xmath203 @xmath204 @xmath205 @xmath206 and @xmath207 respectively . by using these weights
, we can express the hadronization amplitude for the scalar mesons , @xmath138 , in terms of two pseudoscalar mesons .
for instance , we want to reconstruct @xmath3 and @xmath70 from the @xmath12 system in the @xmath208 decay . because of the quark configuration in the parent particle @xmath0 , in this decay the @xmath12 system should be obtained from the hadronization of the @xmath20 pair and the rescattering process for two pseudoscalar mesons , as seen in fig .
[ fig:3 ] , with the weight in eq . .
therefore , for the @xmath209 decay mode we can express the hadronization amplitude with a prefactor @xmath210 and the cabibbo kobayashi
maskawa matrix elements @xmath211 as @xmath212 = c v_{c s } \left ( g_{k^{+ } k^{- } } t_{k^{+ } k^{- } \to \pi ^{+ } \pi ^{- } } \phantom{\frac{1}{2 } } \right .
\notag \\ & \left .
+ g_{k^{0 } \bar{k}^{0 } } t_{k^{0 } \bar{k}^{0 } \to \pi ^{+ } \pi ^{- } } + \frac{1}{3 } \cdot 2 \cdot \frac{1}{2 } g_{\eta \eta } t_{\eta \eta \to \pi ^{+ } \pi ^{- } } \right ) .\end{aligned}\ ] ] in this equation , the decay mode is abbreviated as @xmath213 $ ] , and @xmath214 and @xmath215 are the loop function and scattering amplitude of two pseudoscalar mesons , respectively , whose formulation are given in sec .
[ sec : chua ] .
we have introduced extra factors @xmath216 and @xmath112 for the identical particles @xmath217 .
the former factor @xmath216 comes from the two ways of annihilating the @xmath217 operator in eq . by the @xmath218 state as in the usual manner for effective lagrangians , while the latter one @xmath112 is the symmetry factor for the @xmath217 loop .
the scalar mesons @xmath3 and @xmath5 appear in the rescattering process and exist in the scattering amplitude @xmath215 for two pseudoscalar mesons .
it is important that this is a cabibbo favored process with @xmath211 .
furthermore , since the @xmath20 pair is hadronized , this is sensitive to the component of the strange quark in the scalar mesons . in this study
we assume that @xmath210 is a constant , and hence the hadronization amplitude @xmath138 is a function only of the invariant mass of two pseudoscalar mesons .
here we emphasize that the prefactor @xmath210 should be common to all reactions for scalar meson productions , because in the hadronization the su(3 ) flavor symmetry is reasonable , i.e. , the light quark
antiquark pair @xmath186 hadronizes in the same way regardless of the quark flavor @xmath188 . in this sense
we obtain @xmath219 = c v_{c s } \left ( 1 + g_{k^{+ } k^{- } } t_{k^{+ } k^{- } \to k^{+ } k^{- } } \phantom{\frac{1}{2 } } \right . \notag \\ & \left .
+ g_{k^{0 } \bar{k}^{0 } } t_{k^{0 } \bar{k}^{0 } \to k^{+ } k^{- } } + \frac{1}{3 } \cdot 2 \cdot \frac{1}{2 } g_{\eta \eta } t_{\eta \eta \to k^{+ } k^{- } } \right ) , \end{aligned}\ ] ] for the @xmath220 decay . in this case
we have to take into account the direct production of the two pseudoscalar mesons without rescattering ( the first diagram in fig . [ fig:3 ] ) , which results in the unity in the parentheses . on the other hand , for the @xmath221 decay mode the @xmath222 system should be obtained from the hadronization of @xmath198 and hence this is a cabibbo suppressed decay mode . the hadronization amplitude is expressed as @xmath223 = c v_{c d } \left ( 1 + g_{\pi ^{- } k^{+ } } t_{\pi ^{- } k^{+ } \to \pi ^{- } k^{+ } } \phantom{\frac{1}{\sqrt{2 } } } \right . \notag \\ & \left . \quad
\quad \quad \quad - \frac{1}{\sqrt{2 } } g_{\pi ^{0 } k^{0 } } t_{\pi ^{0 } k^{0 } \to \pi ^{- } k^{+ } } \right ) .
\label{eq : ds_pipi}\end{aligned}\ ] ] in a similar manner we can construct every hadronization amplitude for the scalar meson .
the resulting expressions are as follows : @xmath224 = c v_{c d } \left ( 1 + g_{\pi ^{+ } \pi ^{- } } t_{\pi ^{+ } \pi ^{- } \to \pi ^{+ } \pi ^{- } } \phantom{\frac{1}{2 } } \right .
\notag \\ & + \frac{1}{2 } \cdot 2 \cdot \frac{1}{2 } g_{\pi ^{0 } \pi ^{0 } } t_{\pi ^{0 } \pi ^{0 } \to \pi ^{+ } \pi ^{- } } + \frac{1}{3 } \cdot 2 \cdot \frac{1}{2 } g_{\eta \eta } t_{\eta \eta \to \pi ^{+ } \pi ^{- } } \notag \\ & \left . \ ! \ ! \ ! \ ! \ ! \ ! \phantom{\frac{1}{2 } } + g_{k^{0 } \bar{k}^{0 } } t_{k^{0 } \bar{k}^{0 } \to \pi ^{+ } \pi ^{- } } \right ) , \end{aligned}\ ] ] @xmath225 = c v_{c d } \left ( - \sqrt{\frac{2}{3 } } - \sqrt{\frac{2}{3 } } g_{\pi ^{0 } \eta } t_{\pi ^{0 } \eta \to \pi ^{0 } \eta } \right . \notag \\ & \left .
\phantom{\sqrt{\frac{2}{3 } } } \quad \quad \quad \quad + g_{k^{0 } \bar{k}^{0 } } t_{k^{0 } \bar{k}^{0 } \to \pi ^{0 } \eta } \right ) , \end{aligned}\ ] ] @xmath226 = c v_{c d } \left ( g_{\pi ^{+ } \pi ^{- } } t_{\pi ^{+ } \pi ^{- } \to k^{+ } k^{- } } \phantom{\sqrt{\frac{1}{2 } } } \right .
\notag \\ & + \frac{1}{2 } \cdot 2 \cdot \frac{1}{2 } g_{\pi ^{0 } \pi ^{0 } } t_{\pi ^{0 } \pi ^{0 } \to k^{+ } k^{- } } + \frac{1}{3 } \cdot 2 \cdot \frac{1}{2 } g_{\eta \eta } t_{\eta \eta \to k^{+ } k^{- } } \notag \\ & \left . \
! - \sqrt{\frac{2}{3 } } g_{\pi ^{0 } \eta } t_{\pi ^{0 } \eta \to k^{+ } k^{- } } + g_{k^{0 } \bar{k}^{0 } } t_{k^{0 } \bar{k}^{0 } \to k^{+ } k^{- } } \right ) , \end{aligned}\ ] ] @xmath227 = c v_{c s } \left ( 1 + g_{\pi ^{+ } k^{- } } t_{\pi ^{+ } k^{- } \to \pi ^{+ } k^{- } } \phantom{\frac{1}{\sqrt{2 } } } \right .
\notag \\ & \left . \quad
\quad \quad \quad - \frac{1}{\sqrt{2 } } g_{\pi ^{0 } \bar{k}^{0 } } t_{\pi ^{0 } \bar{k}^{0 } \to \pi ^{+ } k^{- } } \right ) , \end{aligned}\ ] ] @xmath228 = c v_{c d } \left ( \frac{2}{\sqrt{3 } } + \frac{2}{\sqrt{3 } } g_{\pi ^{- } \eta } t_{\pi ^{- } \eta
\to \pi ^{- } \eta } \right . \notag \\ & \left .
\phantom{\frac{2}{\sqrt{3 } } } \quad \quad \quad \quad + g_{k^{0 } k^{- } } t_{k^{0 } k^{- } \to \pi ^{- } \eta } \right ) , \end{aligned}\ ] ] @xmath229 = c v_{c d } \left ( 1 + \frac{2}{\sqrt{3 } } g_{\pi ^{- } \eta } t_{\pi ^{- } \eta \to k^{0 } k^{- } } \right .
\notag \\ & \left .
\phantom{\frac{2}{\sqrt{3 } } } \quad \quad \quad \quad + g_{k^{0 } k^{- } } t_{k^{0 } k^{- } \to k^{0 } k^{- } } \right ) , \end{aligned}\ ] ] @xmath230 = c v_{c s } \left ( 1 + \frac{1}{\sqrt{2 } } g_{\pi ^{0 } k^{- } } t_{\pi ^{0 } k^{- } \to \pi ^{- } \bar{k}^{0 } } \right . \notag \\ & \left .
\phantom{\frac{1}{\sqrt{2 } } } \quad \quad \quad \quad + g_{\pi ^{- } \bar{k}^{0 } } t_{\pi ^{- } \bar{k}^{0 } \to \pi ^{- } \bar{k}^{0 } } \right ) .\end{aligned}\ ] ] the hadronization amplitudes @xmath231 $ ] , @xmath232 $ ] , and @xmath233 $ ] are further simplified by using the isospin symmetry as @xmath234 = c v_{c d } a_{\pi k } , \ ] ] @xmath235 = v_{\rm had}^{(s ) } [ d^{0 } , \ , \pi ^{- } \bar{k}^{0 } ] = c v_{c s } a_{\pi k } , \ ] ] where @xmath236 is a function of the invariant mass of two pseudoscalar mesons and is defined with the scattering amplitude in the isospin basis as @xmath237 in a similar manner , we simplify the hadronization amplitudes @xmath238 $ ] , @xmath239 $ ] , and @xmath240 $ ] as @xmath241 = - \frac{1}{\sqrt{2 } } v_{\rm had}^{(s ) } [ d^{0 } , \ , \pi ^{- } \eta ] = - \sqrt{\frac{2}{3 } } c v_{c d } b_{\pi \eta } , \ ] ] @xmath242 = c v_{c d } b_{k \bar{k } } , \ ] ] with @xmath243 @xmath244 from the above expressions one can easily specify cabibbo favored and suppressed processes for the semileptonic decays into two pseudoscalar mesons , which are listed in table [ tab : mode ] .
finally we note that the use of a constant @xmath210 factor in our approach gets support from the work of ref .
the evaluation of the matrix elements in these processes is difficult and problematic .
there are however some cases where the calculations can be kept under control . for the case of small
recoil , namely when final pseudoscalars move slow , it can be explored in the heavy meson chiral perturbation theory @xcite .
detailed calculations for the case of semileptonic decay are done in @xcite .
there one can see that for large values of the invariant mass of the lepton system the form factors can be calculated and the relevant ones in @xmath245 wave that we need here are smooth in the range of the invariant masses of the pairs of mesons that we use here . to be able to use this behaviour we should prove that in our case the invariant masses of the lepton pair are large , but indeed , it was shown in the study of the semileptonic @xmath26 decays @xcite ( and can be done also here ) that the mass distribution of the lepton pair accumulates at the upper end of the phase space .
there is also another limit , at large recoil , where an approach that combines both hard - scattering and low - energy interactions has been developed is also available @xcite , but this is not the case here .
next we consider processes with the vector mesons in the final state .
as we have already mentioned , hadronization with an extra @xmath35 is unnecessary for the vector mesons . as a consequence
, we can formulate the hadronization amplitude for vector mesons , @xmath246 , in a very simple way . in order to see this
, we consider the semileptonic decay @xmath247 as an example .
the decay process is diagrammatically represented in fig .
[ fig:12](a ) , and the hadronization amplitude @xmath137 can be expressed with a prefactor @xmath248 and the cabibbo
maskawa matrix element @xmath211 as @xmath249 = c^{\prime } v_{c s } , \ ] ] where the decay mode is abbreviated as @xmath250 $ ] in the equation . here we emphasize that the prefactor @xmath248 should be common to all reactions for vector meson productions , as in the case of the scalar meson productions , because the su(3 ) flavor symmetry is reasonable in the hadronization , i.e. , the light quark
antiquark pair @xmath186 hadronizes in the same way regardless of the quark flavor @xmath188 .
we further assume that @xmath248 is a constant again .
this formulation is straightforwardly applied to other vector meson productions and we obtain the hadronization amplitude for vector mesons : @xmath251 = c^{\prime } v_{c d } , \ ] ] @xmath252 = c^{\prime } v_{c s } , \ ] ] @xmath253 = - \frac{1}{\sqrt{2 } } c^{\prime } v_{c d } , \ ] ] @xmath254 = \frac{1}{\sqrt{2 } } c^{\prime } v_{c d } , \ ] ] @xmath255 = - c^{\prime } v_{c s } , \ ] ] @xmath256 = c^{\prime } v_{c d } , \ ] ] where we have used @xmath257 , @xmath258 and @xmath259 states in the isospin basis summarized in appendix [ app:1 ] .
we note that these equations clearly indicate cabibbo favored and suppressed processes with the cabibbo
maskawa matrix elements @xmath211 and @xmath260 , respectively . for the scattering amplitude of two pseudoscalar mesons , we employ the so - called chiral unitary approach @xcite , which we briefly explain in this subsection .
in this approach we solve a coupled - channels bethe
salpeter equation in an algebraic form @xmath261 where @xmath262 , @xmath263 , and @xmath264 are channel indices , @xmath245 is the mandelstam variable of the scattering , @xmath94 is the interaction kernel , and @xmath214 is the two - body loop function . for the hadronization in the previous subsection
we need three types of coupled - channels systems : the @xmath265 system , for which we introduce six channels labeled by the indices @xmath266 , @xmath267 , @xmath268 in the order @xmath269 , @xmath270 , @xmath271 , @xmath272 , @xmath273 , and @xmath274 , the @xmath275-@xmath13 system , and the @xmath276-@xmath277 system .
in this study the interaction kernel @xmath278 is taken as the simplest one , that is , the leading - order @xmath245-wave interaction obtained from the chiral perturbation theory .
the interaction kernel for @xmath279 is summarized as @xmath280 where @xmath188 is the pion decay constant .
one must remember that in the chiral unitary approach when calculating @xmath281 one uses the unitary normalization @xmath282 and @xmath283 for identical particles , which allows to use the general formula in coupled channels . at the end
the good normalization of the external particles must be restored and these are the amplitudes that appear to eq . and following ones . for the @xmath284-@xmath13 scattering , the interaction kernel can be written in terms of the interaction kernel for the @xmath265 system shown above ( see the isospin basis summarized in appendix [ app:1 ] ) : @xmath285 @xmath286 @xmath287 for the @xmath288-@xmath277 scattering , the interaction kernel is expressed as @xmath289 , \end{aligned}\ ] ] @xmath290 , \end{aligned}\ ] ] @xmath291 .
\end{aligned}\ ] ] for the loop function @xmath214 , on the other hand , we use the following expression : @xmath292 where @xmath293 and @xmath294 and @xmath295 are masses of pseudoscalar mesons in channel @xmath262 . in this study
we employ a three - dimensional cut - off @xmath296 as @xmath297^{2 } + i 0 } , \label{eq : gloop}\ ] ] in this expression we have performed the @xmath298 integral and @xmath299 and @xmath300 are the on - shell energies . in this framework , with a small number of free parameters we can reproduce experimental observables of meson meson scatterings fairly well . in this study
we take the model parameters of the chiral unitary approach as @xmath301 and @xmath302 , which dynamically generates resonance poles in the complex energy plane : @xmath303 for @xmath3 , @xmath304 for @xmath70 , and @xmath305 for @xmath4 .
the @xmath306 appears as a cusp at the @xmath21 threshold .
now let us calculate the semileptonic decay widths of @xmath11 mesons into scalar and vector mesons .
as we have formulated , we have only one model parameter for scalar and vector meson productions , respectively . namely one can calculate the decay widths of the scalar meson productions with one common parameter @xmath210 , and similarly @xmath248 for the vector meson productions .
first we consider the scalar meson production in sec .
[ sec : scalar ] , and then move to the vector meson production in sec .
[ sec : vector ] . finally in sec .
[ sec : compare ] we compare the two contributions of the mass distributions from the scalar and vector mesons . in order to calculate the branching fractions of the scalar meson productions , we first fix the prefactor constant @xmath210 so as to reproduce the experimental branching fraction which has the smallest experimental error for the process with the @xmath245-wave two pseudoscalar mesons , that is , @xmath307 = ( 2.32 \pm 0.10 ) \times 10^{-3}$ ] . by integrating the differential decay width , or mass distribution , @xmath308 in an appropriate range , in the case of @xmath309 [ @xmath310 , @xmath311
, we find that @xmath312 can reproduce the branching fraction of @xmath313 . with @xmath314 , @xmath271 , and @xmath222 in @xmath245 wave .
we multiply the @xmath222 mass distribution , which is a cabibbo suppressed process , by @xmath315 .
, title="fig:",width=325 ] with @xmath316 , @xmath274 , @xmath271 , and @xmath317 in @xmath245 wave .
we multiply the @xmath12 , @xmath318 , and @xmath271 mass distributions , which are cabibbo suppressed processes , by @xmath315 .
, title="fig:",width=325 ] with @xmath319 , @xmath320 , and @xmath321 in @xmath245 wave .
we multiply the @xmath322 and @xmath320 mass distributions , which are cabibbo suppressed processes , by @xmath315 .
, title="fig:",width=325 ] by using the common prefactor @xmath312 , we can calculate the mass distributions of two pseudoscalar mesons in @xmath245 wave for all scalar meson modes , which are plotted in figs .
[ fig : dg4_ds ] , [ fig : dg4_dp ] , and [ fig : dg4_dz ] for @xmath0 , @xmath1 , and @xmath2 semileptonic decays , respectively .
we show the mass distributions with the lepton flavor @xmath323 ; the contribution from @xmath324 is almost the same as that from @xmath323 in each meson meson mode due to the small lepton masses . in each figure
we multiply the mass distributions which are cabibbo suppressed processes by @xmath315 so that we can easily compare the shape of the mass distributions .
as one can see , the largest value of the mass distribution @xmath325 is obtained in the @xmath326 process , in which we can see a clear @xmath70 peak .
it is interesting to note that in the @xmath327 process we find a clear @xmath5 signal while the @xmath3 contribution is negligible , which strongly indicates a substantial fraction of the strange quarks in the @xmath5 meson , as we will discuss later . for the @xmath0 semileptonic decay
we also observe a rapid enhancement of the @xmath328 mass distribution at the threshold , as a tail of the @xmath70 contribution , although its height is much smaller than the @xmath329 peak .
for the @xmath1 and @xmath2 semileptonic decays , we can see the @xmath309 and @xmath321 as cabibbo favored processes , respectively .
we note that the @xmath309 and @xmath321 mass distributions are almost the same due to isospin symmetry .
it is interesting to see that the shape of the @xmath309 and @xmath321 mass distributions is determined by , in addition to the @xmath4 resonance , the kinetic factor of the squared decay amplitude .
namely , we have the matrix element of eq .
that is roughly proportional to @xmath330 and this momentum gets bigger the smaller the meson meson invariant mass .
this kinetic factor of the squared decay amplitude affects the @xmath12 distribution in the @xmath1 semileptonic decay in a similar manner , and also provides more weight at low invariant masses for the shape for @xmath13 in figs .
[ fig : dg4_dp ] and [ fig : dg4_dz ] than the @xmath274 distributions in the @xmath331 decay evaluated in ref .
the @xmath13 mass distributions in figs .
[ fig : dg4_dp ] and [ fig : dg4_dz ] of the @xmath1 and @xmath2 decays show peaks corresponding to @xmath332 , but its peak is not high compared to the @xmath5 peak in the @xmath12 mass distribution of the @xmath0 decay since they are obtained in cabibbo suppressed processes .
the @xmath333 decay is cabibbo suppressed and it has a large contribution from the @xmath3 formation and a small one of the @xmath5 , similar to what is found in the @xmath334 decay in ref .
a different way to account for the @xmath335 distribution is by means of dispersion relations , as used in ref .
@xcite in the semileptonic decay of @xmath26 , where the @xmath12 @xmath245-wave distribution has a shape similar to ours .
invariant mass distribution for the semileptonic decay @xmath336 .
the theoretical calculation is folded with the size of experimental bins , @xmath337 .
the experimental data are taken from ref .
@xcite and are scaled so that the fitted breit wigner distribution ( dashed line ) reproduces the branching fraction of @xmath338 = 0.2 \%$ ] by the particle data group ( see table [ tab : br ] ) .
, title="fig:",width=325 ] the theoretical @xmath12 mass distribution of the semileptonic decay @xmath339 is compared with the experimental data @xcite in fig . [
fig : hist ] .
we note that we plot the figure in unit of @xmath340 , not in arbitrary units .
the theoretical mass distribution is folded with @xmath337 spans since the experimental data are collected in bins of @xmath337 . the experimental data , on the other hand , are scaled so that the fitted breit wigner distribution reproduces the branching fraction of @xmath341 = 0.2 \%$ ] @xcite .
the mass and width of the breit wigner distribution are fixed as @xmath342 and @xmath343 , respectively , taken from ref .
@xcite . in fig .
[ fig : hist ] we can see a qualitative correspondence between the theoretical and experimental signals of @xmath5 . we emphasize that , both in experimental and theoretical results , the @xmath12 mass distribution shows a clear @xmath5 signal while the @xmath3 contribution is negligible .
this strongly indicates that the @xmath5 has a substantial fraction of the strange quarks while the @xmath3 has a negligible strange quark component .
it is interesting to recall that the appearance of the @xmath5 in the case one has a hadronized @xmath20 component at the end , and no signal of the @xmath3 , is also observed in @xmath19 and @xmath344 decays in refs .
the explanation of this feature along the lines used in the present work was given in ref .
however , although the peak height of the @xmath5 is very similar , the breit wigner fit would provide larger branching fraction @xmath345 $ ] than the theoretical one . actually ,
integrating the theoretical mass distribution in the range [ @xmath346 , @xmath347 , we obtain the branching fraction @xmath348 = 5.10 \times 10^{-4}$ ] , which is about four times smaller than the experimental value @xmath349 . actually , in the experimental analysis of ref .
@xcite different sources of background are considered that make up for the lower mass region of the distribution .
the width of the @xmath5 extracted in the analysis of ref .
@xcite is @xmath350 , which is large compared to most experiments @xcite , including the lhcb experiment of @xcite , although the admitted uncertainties are also large .
one should also take into account that , while a breit wigner distribution for the @xmath5 is used in the analysis of ref .
@xcite , the large coupling of the resonance to @xmath21 requires a flatte form that brings down fast the @xmath12 mass distribution above the @xmath21 threshold .
our normalization in fig .
[ fig : hist ] is done with the central value of the @xmath351 $ ] and no extra uncertainties from this branching fraction are considered .
yet , we find instructive to do an exercise , adding to our results a `` background '' of @xmath352 from different sources that our calculation does not take into account , and then our signal for the @xmath5 has a good agreement with the peak of the experimental distribution . as mentioned above , the value extracted in @xcite for the @xmath5 signal is tied to the assumptions made , including parts of the background that lead to a very large width of the resonance , assuming a breit wigner shape , etc .
actually , in a more recent paper @xcite the same cleo data of @xcite are reanalyzed taking a band of @xmath5 masses within @xmath353 of @xmath354 and assuming a flatte form of the resonance and a rate for @xmath355 = ( 0.13 \pm 0.02 \pm 0.01 ) \%$ ] is obtained .
this value is about a factor of two smaller than the one reported in @xcite and more in agreement with our results .
decay mode followed by @xmath356 , with @xmath357^{2}$ ] .
the experimental data are taken from ref .
the experimental points should be rescaled dividing by about a factor of two if the absolute rate for the @xmath5 production of the reanalysis of ref .
@xcite were used .
, title="fig:",width=325 ] next we consider the differential decay width with respect to the squared momentum transfer @xmath358 , which coincides with the squared invariant mass of the lepton pair : @xmath359^{2}$ ] . the differential decay width for the scalar meson production
is expressed as @xmath360 this differential decay width was experimentally observed in ref .
@xcite for the @xmath361 decay mode followed by @xmath362 . in this study
we compare our theoretical value for this decay mode with the experimental data in fig .
[ fig : dg4dq2 ] .
the range of the integral for @xmath363 is [ @xmath346 , @xmath347 .
as one can see , we can to some extent reproduce the shape of the differential decay width @xmath364 in experiment , but the absolute value of the theoretical calculation is several times smaller than the experimental one .
this can be , as we have explained , solved by introducing background contributions when extracting the amount of the @xmath5 signal from experimental data .
actually , as we have commented before , the reanalysis of @xcite leads to absolute values of the rate for the @xmath5 production about a factor of two smaller , and again if we scale the @xmath358 distribution of in fig .
[ fig : dg4dq2 ] by the factor the agreement is much better .
8.6cm@lccc + mode & range of @xmath363 [ gev ] & @xmath323 & @xmath365 + @xmath12 & [ 0.9 , 1.0 ] & @xmath366 & @xmath367 + @xmath271 & [ @xmath368 , 1.2 ] & @xmath369 & @xmath370 + @xmath222 & [ @xmath310 , 1.0 ] & @xmath371 & @xmath372 + + + mode & range of @xmath363 [ gev ] & @xmath323 & @xmath365 + @xmath12 & [ @xmath373 , 1.0 ] & @xmath374 & @xmath375 + @xmath274 & [ @xmath376 , 1.1 ] & @xmath377 & @xmath378 + @xmath271 & [ @xmath368 , 1.2 ] & @xmath379 & @xmath380 + @xmath309 & [ @xmath310 , 1.0 ] & @xmath381 & @xmath382 + + + mode & range of @xmath363 [ gev ] & @xmath323 & @xmath365 + @xmath322 & [ @xmath376 , 1.1 ] & @xmath383 & @xmath384 + @xmath320 & [ @xmath368 , 1.2 ] & @xmath385 & @xmath386 + @xmath321 & [ @xmath310 , 1.0 ] & @xmath387 & @xmath388 + moreover , integrating the mass distributions we calculate the branching fractions of the semileptonic @xmath11 mesons into two pseudoscalar mesons in @xmath245 wave , which are listed in table [ tab : g4 ] .
we note that the branching fraction @xmath389= 2.32 \times 10^{-3}$ ] is used as an input to fix the common constant , @xmath312 . among the listed values
, we can compare the theoretical and experimental values of the branching fraction @xmath390 $ ] .
namely , in ref .
@xcite this branching fraction is obtained as @xmath391 of the total @xmath392 , which is dominated by the @xmath32 vector meson .
this indicates , together with the branching fraction @xmath393 , we can estimate @xmath394 = ( 5.5^{+3.1}_{-2.1 } ) \times 10^{-5}$ ] .
theoretically this is @xmath369 .
although our value overestimates the mean value of the experimental data , it is still in @xmath395 errors of the experimental value .
8.6cm@lcc + mode & @xmath323 & @xmath365 + @xmath32 & @xmath396 & @xmath397 + @xmath398 & @xmath399 & @xmath400 + + + mode & @xmath323 & @xmath365 + @xmath401 & @xmath402 & @xmath403 + @xmath404 & @xmath405 & @xmath406 + @xmath8 & @xmath407 & @xmath408 + + + mode & @xmath323 & @xmath365 + @xmath409 & @xmath410 & @xmath411 + @xmath412 & @xmath413 & @xmath414 + let us move to the vector meson productions in the semileptonic @xmath11 decays .
for the vector mesons we fix the common prefactor @xmath248 so as to reproduce the 10 available experimental branching fractions listed in table [ tab : br ] . from the best fit
we obtain the value @xmath415 , which gives @xmath416 .
the theoretical values of the branching fractions are listed in table [ tab : g3 ] and are compared with the experimental data in fig .
[ fig : vec ] , where we plot the ratio of the experimental to theoretical branching fractions .
we calculate the experimental branching fraction of the @xmath417 ( @xmath323 and @xmath97 ) process by dividing the value in table [ tab : br ] by the branching fraction @xmath418 = 2/3 $ ] , which is obtained with isospin symmetry . as one can see from fig .
[ fig : vec ] , the experimental values are reproduced well solely by the model parameter @xmath248 with @xmath419
. decays into vector mesons . , title="fig:",width=325 ] decay mode followed by @xmath420 , with @xmath357^{2}$ ] .
the experimental data are taken from ref .
the theoretical value is multiplied by the branching fraction of the @xmath32 meson to @xmath271 , @xmath421 = 48.9 \%$ ] @xcite .
, title="fig:",width=325 ] next for the @xmath393 decay mode we consider the differential decay width with respect to the squared momentum transfer @xmath358 , which coincides with the squared invariant mass of the lepton pair : @xmath422^{2}$ ] .
this differential decay width was already measured in an experiment @xcite for the @xmath423 decay mode . in a similar manner to the previous case ,
the differential decay width for the vector meson production is expressed as @xmath424 in fig .
[ fig : dg3dq2 ] we compare our result for this reaction with the experimental one .
as one can see , our theoretical result reproduces the experimental value of the differential decay width quantitatively well .
this means that our method to calculate the semileptonic decays of @xmath11 mesons is good enough to describe the decays into vector mesons . in this study
we have not evaluated the @xmath425 decay .
this decay proceeds like the @xmath426 decay that we have evaluated and one has a @xmath200 at the end . since the @xmath192 is @xmath20 then this is forbidden in our approach , at the tree level that we have considered for the vector production .
experimentally , this rate is @xmath57 .
this is an upper bound about @xmath427 times smaller than the rate of the omega production that we have evaluated .
we do not want to go beyond , but can give some idea on how a finite rate could be obtained in our approach .
for this one would have to hadronize the @xmath200 into a @xmath272 , then have a loop for @xmath272 propagation in @xmath428-wave and finally have the @xmath272 couple to the @xmath192 .
some technical details could be borrowed from the study of @xmath429 decay studied in @xcite but one can get an indication that the rate should be rather small by simply noting that the hadronization to meson meson pairs has a reduction factor , as one can see by comparing for instance @xmath430 production with @xmath258 production @xcite . on the other hand ,
the coupling of @xmath192 to @xmath272 is intrinsically small , as one can see from the @xmath431 partial decay width of this channel [ comparatively the @xmath432 partial decay width to the @xmath433 channel would be about @xmath434 for a pion with the same momentum as the kaon in the @xmath192 decay ] .
there are other factors to consider , but this can give us a feeling that the rate could be some orders of magnitude smaller than for omega production .
finally we compare the mass distributions of the two pseudoscalar mesons in @xmath245- and @xmath428-wave contributions . in the present approach
the @xmath245-wave part comes from the rescattering of two pseudoscalar mesons including the scalar meson contribution , while the @xmath428-wave one from the decay of a vector meson . in this study
we consider three decay modes : @xmath72 , @xmath392 , and @xmath435 .
the @xmath436 decay mode would have a large @xmath428-wave contribution from @xmath437 , but we do not consider this decay mode since it is a cabibbo suppressed process .
first we consider the @xmath72 decay mode .
this is a specially clean mode , since it does not have vector meson contributions and is dominated by the @xmath245-wave part .
namely , while the @xmath12 can come from a scalar meson , the primary quark
antiquark pair in the semileptonic @xmath0 decay is @xmath20 , which is isospin @xmath438 and hence the @xmath437 can not contribute to the @xmath12 mass distribution .
the primary @xmath20 can be @xmath32 , but it decays dominantly to @xmath21 and the @xmath439 decay is negligible .
this fact enables us to observe the scalar meson peak without a contamination from vector meson decays and discuss the quark configuration in the @xmath5 resonance as in sec .
[ sec : scalar ] .
invariant mass distribution for the semileptonic decay @xmath392 both in @xmath245 and @xmath428 waves .
, title="fig:",width=325 ] next let us consider the @xmath392 decay mode .
as we have seen , the @xmath271 mass distribution in @xmath245 wave is a consequence of the @xmath5 tail .
however , its contribution should be largely contaminated by the @xmath440 in @xmath428 wave , which has a larger branching fraction than the @xmath441 in the semileptonic decay . in order to see this
, we calculate the @xmath428-wave @xmath271 mass distribution for @xmath392 , which can be formulated as @xmath442 } { [ m_{\rm inv}^{(hh ) } ] ^{2 } - m_{v}^{2 } + i m_{v } \gamma _ { v } ( m_{\rm inv}^{(hh ) } ) } , \ ] ] where @xmath146 is the vector meson mass and the energy dependent decay width @xmath443 is defined as @xmath444 @xmath445^{2 } , \ , m_{h}^{2 } , \ , m_{h}^{\prime 2 } ) } { 2 m_{\rm inv}^{(hh ) } } , \ ] ] @xmath446 for the @xmath32 meson we take @xmath447 and @xmath448 = 0.489 $ ] @xcite . the numerical result for the @xmath449 mass distribution is shown in fig .
[ fig : kpkm ] together with the @xmath450 . from the figure , compared to the @xmath451 contribution we can not find any significant @xmath452 contribution , which was already noted in the experimental mass distribution in ref . @xcite .
nevertheless , we emphasize that the @xmath452 fraction of the semileptonic @xmath0 decay is large enough to be extracted @xcite . actually in ref .
@xcite they extracted the @xmath452 fraction by analysing the interference between the @xmath245- and @xmath428-wave contributions .
this fact , and the qualitative reproduction of the branching fractions in our model , implies that the @xmath5 resonance couples to the @xmath21 channel with a certain strength , which can be translated into the @xmath21 component in @xmath70 , in a similar manner to the @xmath453 component in @xmath454 @xcite , in terms of the compositeness @xcite .
anyway , in order to conclude the structure of the @xmath5 more clearly , it is important to reduce the experimental errors on the @xmath452 .
invariant mass distribution for the semileptonic decay @xmath455 both in @xmath245 and @xmath428 waves .
, title="fig:",width=325 ] finally we consider the @xmath455 decay mode . in this mode
the @xmath456 from the @xmath457 and the @xmath458 from the @xmath459 are competing with each other . in a similar manner to the @xmath392 case
, we calculate the mass distribution also for the @xmath428-wave @xmath309 contribution @xmath460 with @xmath461 @xcite , and the result is shown in fig .
[ fig : ppkm ] . as one can see , thanks to the width of @xmath462 for the @xmath9 , the @xmath245-wave component can dominate the mass distribution below @xmath463 .
we note that we would obtain an almost similar result for the @xmath464 decay mode due to isospin symmetry . as to the theoretical uncertainties , we can play a bit with the cut - offs used to regularize the loops , such that the masses of the states do not change appreciably .
this exercise has been done a number of times and given us the feeling that within our models the uncertainties are below @xmath465 .
for the case of scalar production where we have a range of invariant masses and rely upon a constant production vertex @xmath210 , the changes with the invariant mass in the primary form factors , prior to the final state interaction of the mesons , as found in @xcite , would add some extra uncertainty . in total
it would be fair to accept about @xmath466 uncertainties in this case in the limited range of energies that we move .
in this study we have discussed the semileptonic decays of @xmath11 mesons into light scalar and vector mesons . for the scalar meson production ,
we have formulated the semileptonic decay as the combination of two parts .
one is the weak decay of the charm quark and the emission of a lepton pair via the @xmath100 boson .
the other is a simple hadronization of light @xmath34 pair plus an extra @xmath35 from vacuum into two pseudoscalar mesons after the @xmath100 boson emission , so as to generate the scalar mesons dynamically in the meson meson final state interaction .
the hadronization naturally gives the weight of each pair of pseudoscalar mesons in the decay process , which governs which scalar meson appears in the decay mode .
for the vector mesons , on the other hand , we have not considered the hadronization with an extra @xmath35 and have directly used the light @xmath34 pair after the @xmath100 boson emission as a weight for the vector mesons , which are expected to be genuinely @xmath34 states .
we note that we can specify flavors of quarks contained in the final state scalar and vector mesons by considering cabibbo favored and suppressed processes .
in addition , since the leptons interact only weakly , the semileptonic decay of the heavy meson to two light mesons @xmath69 brings a suitable condition to measure effects of the final state interaction of the two light mesons . in our model of the semileptonic decay ,
the production yields of the scalar and vector mesons are respectively determined solely by constant prefactors @xmath210 and @xmath248 as model parameters .
fixing @xmath210 from the branching fraction of the @xmath467 decay , we have calculated branching fractions of scalar meson productions .
we have qualitatively reproduced the experimental value of the branching fractions of @xmath468 and @xmath469 decay modes .
some deviations of these branching fractions compared to the experimental values can be explained by taking into account the background of the mass distribution for the @xmath12 case and by the large experimental error for the @xmath271 case . for the vector mesons ,
we have determined the constant @xmath248 so as to fit our numerical values to the available experimental values of the branching fractions , and we have reproduced the experimental values at a quantitative level .
we also compared the mass distributions of the two pseudoscalar mesons in @xmath245- and @xmath428-wave contributions , which come from decays of the scalar and vector mesons , respectively .
we have found that the cabibbo favored decay mode @xmath470 followed by @xmath362 and @xmath271 is of special interest . for the @xmath471 mode , we have found that there is no @xmath428-wave contamination from @xmath7 decay and hence it should be dominated by the @xmath245-wave part .
then , we have confirmed the experimental fact that the @xmath12 mass distribution shows a clear @xmath70 signal while the @xmath3 contribution is negligible .
this strongly indicates that the @xmath5 has a substantial fraction of the strange quarks while the @xmath3 has a negligible strange quark component . for the @xmath472 mode ,
on the other hand , the @xmath452 contribution is highly contaminated by the @xmath473 decay in @xmath428 wave .
nevertheless , the @xmath450 fraction of the semileptonic @xmath0 decay is large enough to be extracted experimentally , which implies that the @xmath5 resonance couples to the @xmath21 channel with a certain strength and hence implies a certain amount of the @xmath21 component in @xmath5 .
we appreciate information from s. stone and x. w. kang .
we acknowledge the support by open partnership joint projects of jsps bilateral joint research projects . this work is partly supported by grants - in - aid for scientific research from mext and jsps ( no .
15k17649 and no .
15j06538 ) , the spanish ministerio de economia y competitividad and european feder funds under the contract number fis2011 - 28853-c02 - 01 and fis2011 - 28853-c02 - 02 , and the generalitat valenciana in the program prometeo ii-2014/068 .
we acknowledge the support of the european community - research infrastructure integrating activity study of strongly interacting matter ( acronym hadronphysics3 , grant agreement n. 283286 ) under the seventh framework program of the eu .
we are deeply grateful to the yukawa institute for theoretical physics , kyoto university , where this work was initiated during the yitp workshop yitp - t-14 - 03 on `` hadrons and hadron interactions in qcd '' .
in this appendix we summarize conventions used in this study . in this article
the metric in four - dimensional minkowski space is @xmath474 and the einstein summation convention is used unless explicitly mentioned .
the scalar product of two vectors @xmath475 and @xmath476 is represented as @xmath477 . as the positive and negative energy solutions of the dirac equation
, we express the dirac spinors respectively as @xmath478 and @xmath479 , where @xmath480 is three - momentum of the field and @xmath245 represents its spin .
the dirac spinors are normalized as follows : @xmath481 with @xmath482 and @xmath483 , and hence we have @xmath484 where @xmath485 is the mass of the field , @xmath486 with @xmath487 being the dirac @xmath113 matrices , and @xmath488 is the on - shell four - momentum of the solution .
the identities for the dirac matrices used in this study are summarized as follows : @xmath489 @xmath490 = 4 ( g^{\mu \nu } g^{\rho \sigma } - g^{\mu \rho } g^{\nu \sigma } + g^{\mu \sigma } g^{\nu \rho } ) , \ ] ] @xmath491 = - 4 i \epsilon ^{\mu \nu \rho \sigma } , \ ] ] @xmath492 = \trace \left [ \gamma _ { 5 } \gamma ^{\mu } \gamma ^{\nu } \gamma ^{\rho } \right ] = 0 , \ ] ] where @xmath493 and @xmath494 is the levi - civita symbol with the normalization @xmath495 . the levi - civita symbol satisfies the following identity @xmath496 in terms of the isospin states @xmath497 , the phase convention for pseudoscalar mesons is given by @xmath498 while other pseudoscalar mesons are represented without phase factors . as a result , we can translate the physical two - pseudoscalar meson states into the isospin basis , which we specify as @xmath499 , as @xmath500 @xmath501 @xmath502 @xmath503 @xmath504 @xmath505 the @xmath508 coupling is expressed as @xmath509 with @xmath510 being the coupling constant of the weak interaction , and the @xmath511 coupling as @xmath512 where @xmath513 is the cabibbo kobayashi
maskawa matrix elements for the transition from the charm to light quark @xmath110 .
the @xmath100 boson propagator with four - momentum @xmath514 is written as @xmath515 with the mass of the @xmath100 boson @xmath516 .
the coupling constant @xmath517 and the mass of the @xmath100 boson @xmath516 are related to the fermi coupling constant @xmath101 as @xmath518 in this article we use the following values for physical constants .
the fermi coupling constant : @xmath519 .
the cabibbo kobayashi
maskawa matrix elements : @xmath520 and @xmath521 .
the masses of heavy mesons : @xmath522 , @xmath523 , and @xmath524 .
isospin symmetric masses are used for the light mesons : @xmath525 , @xmath526 , and @xmath527 for the pseudoscalar mesons , and @xmath528 , @xmath529 , @xmath530 , and @xmath531 for the vector mesons .
the masses of the leptons : @xmath532 , @xmath533 , and @xmath534 .
k. a. olive _ et al .
_ [ particle data group collaboration ] , chin .
c * 38 * , 090001 ( 2014 ) .
r. aaij _ et al . _
[ lhcb collaboration ] , phys .
b * 698 * , 115 ( 2011 ) .
j. li _ et al .
_ [ belle collaboration ] , phys .
lett . * 106 * , 121802 ( 2011 ) .
t. aaltonen _ et al . _
[ cdf collaboration ] , phys .
d * 84 * , 052012 ( 2011 ) . v. m. abazov _ et al . _
[ d0 collaboration ] , phys .
d * 85 * , 011103 ( 2012 ) .
r. aaij _ et al . _
[ lhcb collaboration ] , phys .
d * 86 * , 052006 ( 2012 ) . w. h. liang and e. oset , phys .
b * 737 * , 70 ( 2014 ) .
j. a. oller and e. oset , nucl .
a * 620 * , 438 ( 1997 ) [ erratum - ibid .
a * 652 * , 407 ( 1999 ) ] .
n. kaiser , eur .
j. a * 3 * , 307 ( 1998 ) . m. p. locher , v. e. markushin and h. q. zheng , eur .
j. c * 4 * , 317 ( 1998 ) .
j. a. oller , e. oset and j. r. pelaez , phys .
* 80 * , 3452 ( 1998 ) .
j. a. oller , e. oset and j. r. pelaez , phys .
d * 59 * , 074001 ( 1999 ) [ erratum - ibid .
d * 60 * , 099906 ( 1999 ) ] [ erratum - ibid .
d * 75 * , 099903 ( 2007 ) ] .
j. a. oller and e. oset , phys .
d * 60 * , 074023 ( 1999 ) .
j. nieves and e. ruiz arriola , nucl .
a * 679 * , 57 ( 2000 ) .
j. r. pelaez and g. rios , phys .
lett . * 97 * , 242002 ( 2006 ) .
m. albaladejo and j. a. oller , phys . rev .
* 101 * , 252002 ( 2008 ) .
a. martinez torres , l. s. geng , l. r. dai , b. x. sun , e. oset and b. s. zou , phys
. lett .
b * 680 * , 310 ( 2009 ) .
j. j. xie , l. r. dai and e. oset , phys .
b * 742 * , 363 ( 2015 ) .
f. s. navarra , m. nielsen , e. oset and t. sekihara , phys .
d * 92 * , no . 1 , 014031 ( 2015 ) . m. ablikim _ et al .
_ [ bes collaboration ] , eur . phys .
j. c * 47 * , 31 ( 2006 ) .
m. ablikim _ et al . _
[ bes collaboration ] , eur .
j. c * 47 * , 39 ( 2006 ) .
j. m. link _ et al . _ [ focus collaboration ] , phys .
b * 598 * , 33 ( 2004 ) .
j. m. link _ et al . _
[ focus collaboration ] , phys .
b * 637 * , 32 ( 2006 ) .
b. aubert _ et al . _ [ babar collaboration ] , phys .
d * 78 * , 051101 ( 2008 ) .
del amo sanchez _ et al .
_ [ babar collaboration ] , phys . rev .
d * 83 * , 072001 ( 2011 ) .
j. yelton _ et al . _
[ cleo collaboration ] , phys .
d * 80 * , 052007 ( 2009 ) .
k. m. ecklund _ et al . _
[ cleo collaboration ] , phys .
d * 80 * , 052009 ( 2009 ) .
l. martin _ et al . _
[ cleo collaboration ] , phys .
d * 84 * , 012005 ( 2011 ) .
j. yelton _ et al . _
[ cleo collaboration ] , phys .
d * 84 * , 032001 ( 2011 ) s. dobbs _ et al .
_ [ cleo collaboration ] , phys .
lett . * 110 * , no .
13 , 131802 ( 2013 ) .
i. bediaga , f. s. navarra and m. nielsen , phys .
b * 579 * , 59 ( 2004 ) .
h. w. ke , x. q. li and z. t. wei , phys .
d * 80 * , 074030 ( 2009 ) .
n. n. achasov and a. v. kiselev , phys .
d * 86 * , 114010 ( 2012 ) .
a. h. fariborz , r. jora , j. schechter and m. naeem shahid , phys .
d * 84 * , 094024 ( 2011 ) .
a. h. fariborz , r. jora , j. schechter and m. n. shahid , int .
j. mod .
a * 30 * , no .
02 , 1550012 ( 2015 ) .
w. wang and c. d. lu , phys .
d * 82 * , 034016 ( 2010 ) .
g. meissner and w. wang , phys .
b * 730 * , 336 ( 2014 ) .
w. wang and r. l. zhu , phys .
b * 743 * , 467 ( 2015 ) .
u. g. meissner and w. wang , jhep * 1401 * , 107 ( 2014 ) . w. f. wang , h. n. li , w. wang and c. d. lu , phys .
d * 91 * , no .
9 , 094024 ( 2015 ) .
x. w. kang , b. kubis , c. hanhart and u. g. meissner , phys .
d * 89 * , 053015 ( 2014 ) .
m. doring , u. g. meissner and w. wang , jhep * 1310 * , 011 ( 2013 ) .
y. j. shi and w. wang , arxiv:1507.07692 [ hep - ph ] .
a. bramon , a. grau and g. pancheri , phys .
b * 283 * , 416 ( 1992 ) .
a. v. manohar and m. b. wise , `` heavy quark physics , '' camb .
monogr . part .
cosmol .
* 10 * , 1 ( 2000 ) .
r. aaij _ et al . _
[ lhcb collaboration ] , phys .
d * 89 * , no . 9 , 092006 ( 2014 ) .
j. hietala , d. cronin - hennessy , t. pedlar and i. shipsey , phys .
d * 92 * , no . 1 , 012009 ( 2015 ) . j. a. oller , e. oset and j. r. pelaez , phys .
d * 62 * , 114017 ( 2000 ) .
m. bayar , w. h. liang and e. oset , phys .
d * 90 * , no .
11 , 114004 ( 2014 ) .
m. albaladejo , m. nielsen and e. oset , phys .
b * 746 * , 305 ( 2015 ) .
t. sekihara , t. hyodo and d. jido , ptep * 2015 * , no . 6 , 063d04 . | we study the semileptonic decays of @xmath0 , @xmath1 , and @xmath2 mesons into the light scalar mesons [ @xmath3 , @xmath4 , @xmath5 , and @xmath6 and the light vector mesons [ @xmath7 , @xmath8 , @xmath9 , and @xmath10 . with the help of a chiral unitarity approach in coupled channels
, we compute the branching fractions for scalar meson processes of the semileptonic @xmath11 decays in a simple way . using current known values of the branching fractions ,
we make predictions for the branching fractions of the semileptonic decay modes with other scalar and vector mesons .
furthermore , we calculate the @xmath12 , @xmath13 , @xmath14 , and @xmath15 invariant mass distributions in the semileptonic decays of @xmath11 mesons , which will help us clarify the nature of the light scalar mesons . |
let @xmath3 be a connected orientable @xmath4-smooth surface possibly with boundary , and let @xmath5 denote the universal cover of @xmath3 .
a _ ( complex ) projective structure _
@xmath1 on @xmath3 is a @xmath6-structure , where @xmath7 is the riemann sphere . in other words
, it is a maximal atlas of @xmath3 modeled on @xmath8 with transition maps in @xmath9 .
the pair @xmath10 is called _ a projective surface_. as usual , we will often conflate @xmath1 and @xmath10 .
there is an equivalent definition , which we will mostly use in this paper : a projective structure is a pair @xmath11 , where @xmath12 is a @xmath4-smooth locally injective map and @xmath13 is a homomorphism , such that @xmath14 is @xmath15-equivariant , i.e. @xmath16 for all @xmath17 . then @xmath14 is called the _ developing map _ and @xmath15 the _ holonomy ( representation ) _ of the projective structure . on the interior of @xmath5
, @xmath18 is a local homeomorphism , and the restriction of @xmath14 to each boundary component of @xmath5 is a @xmath4-smooth curve . a projective structure @xmath19 is defined up to an isotopy of @xmath3 and the action of an element of @xmath9 , i.e. the post - composition of @xmath14 with @xmath20 and the conjugation of @xmath15 by @xmath21 .
( see @xcite , @xcite . )
the @xmath4-smoothness is required to define a natural topology on the space of all projective structures on @xmath3 in the case that @xmath3 is not closed ( see ( * ? ? ?
* i.1.5 ) ) . in this paper
, we always assume that surfaces are connected and @xmath4-smooth possibly with boundary ( although we do not deform a projective structure ) .
a projective structure @xmath22 on @xmath3 is * admissible * if @xmath14 is an embedding and @xmath15 is an isomorphism onto a schottky group or a quasifuchsian group .
analogously , a simple loop @xmath23 on the projective surface @xmath10 is * admissible * if @xmath24 embeds to @xmath8 by @xmath14 and @xmath25 is loxodromic , where @xmath24 is a lift of @xmath23 to @xmath5 and @xmath26 is the homotopy class of @xmath23 .
a hyperbolic structure is a basic example of a projective structure , since @xmath27 and @xmath28 in a compatible way . besides , every hyperbolic structure on a closed orientable surface is an admissible projective structure .
however , in general , developing maps are not necessarily injective and holonomy representations are not necessarily discrete or faithful ( c.f .
( i ) and ( ii ) below ) . throughout this paper ,
let @xmath0 denote a closed orientable surface of genus at least @xmath29 and @xmath30 denote the universal cover of @xmath0 .
( orientability of @xmath0 is not essential for the mains theorems of this paper , if we consider two - dimensional mobius structures instead of projective structures . )
the following theorem yields a decomposition of an arbitrary projective surface into admissible projective subsurfaces : + * theorem [ i d ] * let @xmath1 be a projective structure on @xmath0 .
then there exists a decomposition of @xmath0 into cylinders and compact connected surfaces of negative euler characteristic , such that the restriction of @xmath1 to each cylinder is an integral flat structure and the restriction to each surface of negative euler characteristic is an admissible projective structure .
+ an _ integral flat structure _ is a basic projective ( actually affine ) structure on a cylinder , which can be obtained by an operation called _ grafting _ ( [ flatcylinders ] ) .
if there is an admissible loop on a projective surface , we can define a grafting along this loop ( see @xcite , @xcite for example ) .
this operation gives another projective structure on the same surface , preserving the orientation and the holonomy representation .
if the admissible loop is circular , i.e. it corresponds to a simple circular arc on @xmath8 via the developing map , then the integral flat structure is exactly the structure that the grafting operation inserts to the projective structure along the loop .
a flat structure on an annulus can be easily decomposed into admissible flat structures .
theorem [ i d ] immediately implies : + * corollary [ pants ] * let @xmath1 be an arbitrary projective structure on @xmath0 . then there exists a decomposition of @xmath0 into pairs of pants and cylinders such that the restriction of @xmath1 to each cylinder is an integral flat structure and the restriction to each pair of pants is an admissible structure . +
this corollary gives the affirmative answer to a question raised by gallo , kapovich , and marden ( @xcite ) .
the authors of @xcite gave necessary and sufficient conditions for a representation @xmath31 to be the holonomy representation of some projective structure on @xmath0 .
the conditions are : ( i ) @xmath32 is a non - elementary subgroup of @xmath9 and ( ii ) @xmath15 lifts to a representation from @xmath33 to @xmath34 . in order to prove the sufficiency , given an arbitrary representation @xmath15 satisfying ( i ) and ( ii )
, they constructed a projective structure on @xmath0 with holonomy @xmath15 in the following way : first , decompose @xmath0 into pairs of pants , @xmath35 , such that @xmath36 is an isomorphism onto a rank - two schottky group for each @xmath37 .
second , construct an admissible projective structure on each @xmath38 with the holonomy representation @xmath36 .
last , glue these structures on the pairs of pants together by inserting projective structures on cylinders between the corresponding boundaries of @xmath38 s , and obtain a desired projective structure .
they asked whether every projective structure on @xmath0 arises from such a schottky pants decomposition .
more specifically , they asked if every projective structure contains an admissible loop , which is answered by : + * theorem [ gc ] * for every projective structure @xmath1 on @xmath0 , there exists an admissible loop on @xmath2 .
+ theorem [ gc ] is weaker than theorem [ i d ] , since the boundary components of each subsurface in theorem [ i d ] are , in particular , admissible loops .
however , the proof of theorem [ gc ] contains the basic ideas for the proof of theorem [ i d ] .
furthermore , theorem [ gc ] addresses the following question about grafting : + * question 1 * ( @xcite ) : assume that two projective structures on @xmath0 have the same orientation and holonomy representation .
can one projective structure be transformed to the other by a sequence of graftings and inverse - graftings ?
+ the grafting and inverse - grafting operations generate an equivalence relation among the projective structures with a given holonomy representation .
question 1 asks if there are exactly two equivalence classes represented by the orientations of the projective structures .
theorem [ gc ] implies that every equivalence class consists of infinitely many projective structures .
ultimately , question 1 aims to characterize the collection of projective structures with the given holonomy representation ( see @xcite , @xcite ) . in the special case
that the holonomy representation is an isomorphism onto a quasifuchsian group , the characterization is given by goldman , using grafting , and the answer to question 1 is affirmative ( ( * ? ? ?
* theorem c ) ) .
also , we may compare this discussion about question 1 with the fact that dehn twists generate the mapping class group of @xmath0 and yield a finite presentation of this group . the _ holonomy map _
@xmath39 is a projection given by @xmath40 , where @xmath41 is the space of all projective structures on @xmath0 and @xmath42 is the representation variety of homomorphisms from @xmath33 to @xmath9 .
this map is _ not _ a covering map onto its image ( @xcite ) , which makes problems in this area difficult .
one may ask the above questions in the case of other @xmath43-structures as well ( c.f .
@xcite ) . in particular
, s. choi gave a canonical decomposition of real projective structures , i.e. @xmath44-structures , analogous to the one given by theorem [ i d ] ( see @xcite ) .
+ * an outline of the proofs : * for a given projective surface @xmath2 , there is a corresponding pair @xmath45 of a marked hyperbolic structure @xmath46 on @xmath0 and a measured geodesic lamination @xmath47 on @xmath48 ( see thurston s coordinates in [ thp ] ) .
a periodic leaf of @xmath49 corresponds to a continuous family of admissible loops on @xmath2 . for each irrational minimal sublamination of @xmath49 , consider a standard sequence @xmath50 of simple geodesic loops on @xmath48 that approximates @xmath49 ( [ appro ] ) .
then we will show that @xmath51 is admissible for all sufficiently large @xmath37 .
these admissible loops yield theorem [ gc ] . taking a disjoint union of such admissible loops ,
we construct a multi - loop on @xmath2 well approximating the entire @xmath52 , the union of leaves of @xmath49 .
we will show that the corresponding multi - loop on @xmath2 achieves the desired admissible decomposition described in theorem [ i d ] .
every measured lamination on @xmath53 induces a continuous map from @xmath53 to @xmath54 , called a _ bending map _ ( [ bending ] ) . via the bending map
, the measure lamination corresponds to a projective structure on @xmath55 ( [ thp ] ) .
our proofs are based on the fact that injective quasiisometric ( bending ) maps correspond to admissible projective structures on @xmath55 ( [ ibm ] ) . in order to show that @xmath51 is admissible
, we take the _ total lift _
@xmath56 of @xmath47 to @xmath53 ( [ dt ] ) and a lift @xmath57 of @xmath51 to @xmath53 .
consider the sublamination of @xmath56 that consists of the leaves of @xmath56 intersecting @xmath57 , so that the structure on @xmath58 embeds into the projective structure on @xmath55 corresponding to the sublamination .
in other words , @xmath56 and the sublamination coincide in a sufficiently small neighborhood of @xmath58 , and therefore , the sublamination is sufficient to capture the structure on @xmath58 and @xmath51
. thus sublamination induces an injective quasiisometric embedding since the bending " is uniformly small ( [ al ] , [ ibm ] ) .
then the sublamination of @xmath56 corresponds to an admissible structure on @xmath55 . since the structure on @xmath57
is embedded in the admissible structure on @xmath55 , therefore @xmath51 is also admissible .
theorem [ i d ] will be proven based on the same idea . +
* acknowledgments .
* i would like to thank my advisor , misha kapovich , for introducing me to complex projective structures and supporting me with his comments and encouragements during this work .
i was partially supported by nsf grants .
@xmath59 : the geodesic segment connecting @xmath60 and @xmath61 , where @xmath60 and @xmath61 are points in a metric space .
+ @xmath62 : the closure of @xmath63 , where @xmath63 is a subset of a topological space .
+ @xmath64 : the interior of @xmath63 .
+ @xmath65 : the convex hull of @xmath66 , where @xmath66 is a subset of a hyperbolic space .
+ @xmath67 : the closed disk of radius @xmath68 centered at @xmath60 in a hyperbolic space .
( for details , see @xcite , @xcite , @xcite , @xcite , @xcite . )
let @xmath3 be a riemannian surface with a constant curvature , possibly with geodesic boundary .
a _ geodesic lamination _ on @xmath3 is a collection of disjoint simple complete geodesics on @xmath3 whose union is a closed subset of @xmath3 .
each geodesic of a geodesic lamination is called a _
leaf_. for a measured lamination @xmath49 on @xmath3 , let @xmath69 denote the union leaves of @xmath49 .
if @xmath70 , then @xmath49 is called a _ foliation_. a _ measured lamination _ is a pair @xmath71 , where @xmath49 is a geodesic lamination , and @xmath72 is a transversal measure of @xmath49 . let a leaf of @xmath47 refer to a leaf of @xmath49 and @xmath73 refer to @xmath74 .
we always assume that @xmath49 is the support of @xmath72 , i.e. if @xmath23 is a leaf of @xmath47 , then @xmath75 for every geodesic segment @xmath76 that transversally intersects @xmath23 .
the _ weight _ of a leaf @xmath23 of @xmath47 is @xmath77 , where @xmath76 varies over all geodesic segments that transversally intersect @xmath23 , i.e. the atomic transversal measure of @xmath23 .
we denote the weight of @xmath23 by @xmath78 . if @xmath23 is an isolated leaf of @xmath47 , then @xmath79 for every geodesic segment @xmath76 on @xmath3 that transversally intersects @xmath80 exactly once at a point on @xmath23 . by convention ,
if a geodesic segment @xmath76 is contained in a leaf of @xmath49 , then @xmath81 . in addition , letting @xmath82 be the end points of @xmath76 , if @xmath60 or @xmath61 is contained in a leaf with positive weight , then the weight does not contribute to the value of @xmath83 , so that @xmath84 where @xmath85 varies over all geodesic segments strictly contained in @xmath76 , i.e. @xmath86 .
recall that @xmath0 is a closed orientable hyperbolic surface
. a measured lamination on @xmath0 is _ minimal _ if it does not contain any proper sublamination .
every measured lamination @xmath47 on @xmath0 uniquely decomposes into a finite number of disjoint minimal laminations of the following two types : a periodic leaf with positive weight ( _ periodic minimal lamination _ ) , and a measured lamination consisting of uncountably many bi - infinite geodesics ( _ irrational minimal lamination _ ) .
let @xmath87 be an irrational minimal sublamination of @xmath47 . then , since each minimal sublamination of @xmath47 forms a closed subset of @xmath0 , @xmath88 is also an open subset of @xmath52 . here
are other properties of @xmath89 : each leaf of @xmath90 is a dense subset of @xmath91 .
if @xmath76 is a geodesic segment on @xmath0 transversally intersecting @xmath90 , then @xmath92 is a cantor set i.e. , a closed , perfect subset with empty interior ( ( * ? ? ?
* corollary 1.7.6 , 1.7.7 ) ) .
therefore , no leaf of @xmath89 has a positive weight .
for every @xmath93 and distinct @xmath94 , let @xmath95 be the open region in @xmath8 bounded by two simple circular arcs connecting @xmath96 and @xmath97 such that the inner angles at the vertices @xmath96 and @xmath97 are equal to @xmath98 . since @xmath95 is embedded in @xmath8 , it is equipped with a canonical projective structure ( whose developing map is the identity map ) .
we call the structure a _ crescent _ of angle @xmath98 ; its projective structure only depends on the choice of @xmath98 .
let @xmath99 be a hyperbolic element of @xmath9 that fixes @xmath96 and @xmath97 .
then @xmath100 , the subgroup of @xmath9 generated by @xmath99 , is an infinite cyclic group acting on @xmath95 freely and property discontinuously . by quotienting @xmath95 by @xmath100
, we obtain a projective structure on a cylinder .
we call the structure a _ flat structure _ of height @xmath98 ; it only depends on @xmath98 and the translation length of @xmath99 . a flat structure of height @xmath98 on a cylinder forms a projective surface , which we call a _ flat cylinder _ of height @xmath98 .
we shall define a crescent and a flat cylinder for arbitrary @xmath101 , generalizing those for @xmath102 . for arbitrary @xmath101 , let @xmath103 .
define @xmath104 by @xmath105 ( i.e. the polar coordinates ) .
then @xmath106 defines a projective structure on @xmath107 , which is a crescent of angle @xmath98 . for @xmath108 ,
let @xmath109 be an automorphism of @xmath110 defined by @xmath111 .
define a homomorphism @xmath112 by @xmath113 for all @xmath114 .
then @xmath106 is @xmath15-equivariant .
quotienting @xmath95 by the action of @xmath115 , we obtain a flat cylinder of height @xmath98 .
a flat cylinder of height @xmath98 is _ integral _ if @xmath98 is a multiple of @xmath116 .
the multiplier is called the _ degree _ of the integral flat cylinder , so that , for all @xmath117 , the cardinality of @xmath118 equals to the degree .
let @xmath119 be the crescent of angle @xmath101 given in the form above , where @xmath120 is the trivial representation .
for each @xmath121 , @xmath106 takes @xmath122 to a straight line on @xmath123 connecting @xmath124 and @xmath125 .
the collection of these lines , @xmath126 , forms a foliation @xmath127 on @xmath95 , which we call the _ canonical foliation _ on @xmath1 .
we also can define the canonical transversal measure @xmath128 of @xmath127 by @xmath129 for all @xmath130 .
note that @xmath131 and @xmath132 bound a crescent of height @xmath133 contained in @xmath1 .
call @xmath134 the _ canonical measured foliation _ on @xmath1 . for each @xmath135
, @xmath136 is orthogonal to ( each leaf of ) @xmath127 in terms of the angles obtained by pulling back the conformal structure on @xmath8 via @xmath106 .
besides , @xmath106 takes @xmath137 to a ( not necessarily simple ) circular arc on @xmath8 .
the collection of these orthogonal lines @xmath138 forms a foliation on @xmath95 , which is dual to @xmath127 . by identifying the points on each leaf of the dual foliation , @xmath95 projects to a line .
since @xmath134 and its dual foliation are invariant under the action of @xmath115 , we obtain the _ canonical foliation _ and its dual foliation on the flat cylinder @xmath139 .
accordingly , the flat cylinder projects to a circle by identifying the points on each leaf of the dual foliation . since @xmath104 continuously extends to @xmath140 , we can compactify a flat cylinder of height @xmath98 to a projective structure on a compact cylinder with boundary . by abusing the notation ,
we call this compactified flat cylinder , also , a flat cylinder of angle @xmath98 . accordingly
, the universal cover of the compactified flat cylinder of angle @xmath98 also is called a crescent of angle @xmath98 .
( for more details , see @xcite , @xcite . )
let @xmath0 be a closed hyperbolic surface and @xmath71 be a measured lamination on @xmath0 without periodic leaves
. then no leaf of @xmath47 has a positive weight .
let @xmath141 be the covering map .
the _ total lift _ of @xmath47 is a measured lamination @xmath142 on @xmath53 , where @xmath143 consists of all the lifts of the leaves of @xmath49 , and @xmath144 is the pull back of @xmath72 , so that @xmath145 is locally isomorphic to @xmath146 via @xmath147 .
then @xmath56 is a @xmath33-invariant measured lamination on @xmath53 .
there is a unique @xmath148-tree dual to @xmath56 constructed in the following way ( if @xmath47 contains periodic leaves , the construction is more complicated ) .
the transversal measure @xmath144 defines a pseudo - metric @xmath149 on @xmath53 , by @xmath150 for all @xmath151 .
since no leaf of @xmath56 has a positive weight , @xmath152 is continuous .
define an equivalence relation on @xmath53 by @xmath153 if and only if @xmath154 .
there are only two types of the equivalence classes : the closure of a complementary region of @xmath155 , and a leaf of @xmath143 that is _ not _ a boundary geodesic of such a complementary region .
let @xmath156 be the quotient of @xmath53 by the equivalence relation , and let @xmath157 be the quotient map .
in particular , if a geodesic in @xmath53 is a leaf of @xmath56 or it is contained in the closure of a complementary region of @xmath155 , then @xmath158 takes this geodesic to a point in @xmath156 .
it turns out that @xmath156 is an @xmath148-tree equipped with a canonical metric @xmath159 induced by @xmath144 : for every @xmath160 , @xmath161 , where @xmath162 are such that @xmath163 and @xmath164 .
if a geodesic in @xmath53 transversally intersects to @xmath56 , then @xmath158 takes this geodesic to a geodesic in @xmath156 .
since the action of @xmath33 on @xmath53 preserves @xmath56 , @xmath165 isometrically acts on @xmath156 .
let @xmath23 be a closed geodesic on @xmath0 transversally intersecting @xmath47 , @xmath166 be a lift of @xmath23 to @xmath53 , and @xmath26 be the homotopy class of @xmath23 in @xmath33 .
then @xmath167 is a geodesic in @xmath156 , and the action of @xmath26 on @xmath156 isometrically translates along @xmath167 by the distance @xmath168 .
let @xmath169 and @xmath170 be open intervals in @xmath148 , and let @xmath66 be a closed subset of @xmath170 . consider a geodesic lamination @xmath49 on @xmath171 that consists of the leaves @xmath172 for all @xmath173 ( i.e. we have @xmath66-worth of horizontal leaves ) .
let @xmath72 be a transversal measure for @xmath49 , and let @xmath47 be the measured lamination @xmath174 .
the pair @xmath175 is called a _ ( euclidian ) flow box_. let @xmath76 be a vertical geodesic in @xmath176 , i.e. @xmath177 for some @xmath178
. the _ height _ of the flow box is @xmath83 , which does not depend on the choice of @xmath60 .
if @xmath47 is a measured lamination on a hyperbolic quadrilateral @xmath179 , and @xmath180 is isomorphic to a flow box , then we also call @xmath180 a ( hyperbolic ) flow box . [ box - nhd ]
let @xmath71 be a measured geodesic lamination on @xmath53 without leaves of positive weight .
let @xmath76 be a geodesic segment contained in a leaf of @xmath47 .
for every @xmath181 , there exists a neighborhood of @xmath76 isomorphic to a flow box of height less than @xmath182 .
let @xmath23 be the leaf of @xmath47 containing @xmath76 .
let @xmath85 be a geodesic segment strictly containing @xmath76 , i.e. @xmath183 and the end points of @xmath76 and @xmath85 are all distinct .
let @xmath184 and @xmath185 be the geodesic segments orthogonal to @xmath85 and passing through the end points of @xmath85 .
let @xmath186 and @xmath187 be the components of @xmath188 , which are the open half planes bounded by @xmath23 .
choose a geodesic @xmath51 in @xmath189 sufficiently close to @xmath23 such that @xmath51 does not intersect @xmath47 transversally .
for @xmath190 , let @xmath191 denote the open hyperbolic quadrilateral bounded by @xmath23 , @xmath184 , @xmath51 , @xmath185 .
let @xmath192 and @xmath193 be the opposite edges of @xmath191 that are contained in @xmath184 and @xmath185 , respectively .
note that the edge of @xmath191 contained in @xmath23 is @xmath85 . first
, suppose that the leaves of @xmath47 contained in @xmath189 do _ not _ accumulate to @xmath23 .
if we choose @xmath51 sufficiently close to @xmath23 , @xmath194 does not intersect leaves of @xmath47 . then @xmath191 is a flow box with the empty lamination .
next , suppose that the leaves of @xmath47 contained in @xmath189 accumulate to @xmath23 .
if @xmath51 is sufficiently close to @xmath23 , then , for every leaf @xmath195 of @xmath47 that intersects @xmath191 , @xmath196 intersects both @xmath192 and @xmath193 . therefore , using the klein model of @xmath53 , one can easily see that @xmath191 is a flow box .
the height of the flow box is @xmath197 .
since no leaf of @xmath47 has a positive wight , we can assume that the height is less than @xmath198 , again by choosing @xmath51 sufficiently close to @xmath23 .
hence , @xmath199 is a flow box neighborhood of @xmath76 with height less than @xmath182 . _
remark : _ under the projection map from @xmath200 to its dual tree , the flow box neighborhood of height less than @xmath182 projects to a geodesic segment of length less than @xmath182 .
( for details , see ( * ? ? ?
* i.4.2.15 ) . )
let @xmath0 be a closed ( orientable ) hyperbolic surface , and @xmath71 be an irrational minimal measured lamination on @xmath0 .
we shall construct a sequence of simple geodesic loops that approximates @xmath52 . pick a leaf @xmath23 of @xmath49 and a point @xmath201 on @xmath23 .
let @xmath76 be a geodesic segment transversally intersecting @xmath23 at @xmath201 .
recall that @xmath202 is a perfect subset of @xmath76 .
in addition , @xmath23 is a dense subset of @xmath52 .
therefore , if we walk along @xmath23 , starting form @xmath201 , we never return to @xmath201 , but we pass by @xmath201 arbitrarily closely .
more precisely , there exists a sequence of points @xmath38 in @xmath203 , such that @xmath204 and @xmath205 , where @xmath192 is a geodesic segment in @xmath23 connecting @xmath201 and @xmath38 and @xmath193 is the geodesic segment on @xmath0 realizing the distance between @xmath201 and @xmath38 on @xmath0 ( figure [ a_i&b_i ] ) . since @xmath205 and @xmath206 for all @xmath37 , by taking a subsequence if necessary , we can assume that @xmath192 and @xmath193 intersect only at their end points , so that @xmath207 is a simple loop on @xmath0 .
homotope this simple loop to a simple geodesic loop @xmath51 on @xmath0 .
it turns out that @xmath50 limits to @xmath52 in the chabauty topology .
besides , @xmath51 transversally intersects @xmath74 ( uncountably many times ) .
let @xmath195 and @xmath208 be geodesics on a complete hyperbolic surface .
assume that @xmath209 , and pick @xmath210 .
we can `` rotate '' @xmath195 to @xmath208 about @xmath147 by a unique angle in @xmath211 $ ] .
more precisely , we do the following : let @xmath212 be a lift of @xmath147 to @xmath53 , and let @xmath213 and @xmath214 be the lifts of @xmath195 and @xmath208 , respectively , to @xmath53 that intersect at @xmath212 .
then , we can indeed rotate @xmath213 to @xmath214 about @xmath212 by a unique angle in @xmath211 $ ] .
let @xmath215 denote this angle , and call it the _ angle _ between @xmath195 and @xmath208 at @xmath147 .
let @xmath90 be a geodesic lamination on the surface that intersects @xmath208 .
the * angle * between @xmath90 and @xmath208 is @xmath216.\ ] ] [ ml1 ] @xmath217 and @xmath218 .
choose @xmath219 , and let @xmath220 be the leaf of @xmath49 that intersects @xmath51 at @xmath221 .
let @xmath222 .
note that @xmath52 is a compact subset of @xmath0 .
therefore , by taking a subsequence if necessary , @xmath223 converges to @xmath224 $ ] .
assume that @xmath225 .
then there exists a sequence @xmath223 converging to @xmath226 with @xmath227 .
the sequence @xmath50 converges to a geodesic that intersects @xmath52 at @xmath60 with angle @xmath98 .
this contradicts the convergence of @xmath50 to @xmath52 .
next , we prove that @xmath228 .
let @xmath229 , which is a simple loop on @xmath0 .
observe that @xmath230 , since @xmath231 and @xmath232 .
therefore , it suffices to show that @xmath233 .
the basic idea is that a geodesic loop realizes the minimal transversal measure among all the loops in the same homotopy class .
let @xmath234 be the homotopy class of @xmath51 .
regarding @xmath51 as a bi - infinite geodesic , let @xmath57 be the lift of @xmath51 , such that @xmath235 translates @xmath53 along the geodesic @xmath57 by @xmath236 .
regarding @xmath51 as a simple closed path , let @xmath237 be a lift of @xmath51 , such that that @xmath237 is contained in @xmath57 .
then @xmath235 identifies the ends of @xmath237 . recall that @xmath158 is the projection from @xmath53 to the @xmath148-tree @xmath156 dual to @xmath56 ( [ dt ] ) .
then @xmath238 is a geodesic in @xmath156 , and the action of @xmath235 on @xmath156 isometrically translates along @xmath239 by @xmath240 .
similarly , regarding @xmath241 as a simple closed path , let @xmath242 be a lift of @xmath241 to @xmath53 , such that the ends of @xmath242 are identified by @xmath235 .
then @xmath243 is a piecewise geodesic path in @xmath156 , and @xmath235 identifies the ends of @xmath243 .
we also have @xmath244 .
the translation length of @xmath235 is equal to or less than the distance between the ends of @xmath243 since the translation length of @xmath235 is @xmath245 .
since @xmath235 translates along @xmath246 , the translation length is @xmath247 .
hence , @xmath233 .
( for details , see @xcite.)[bending ] let @xmath71 be a measured lamination on @xmath53 .
then @xmath47 induces a _ bending map _ @xmath248 by bending @xmath53 inside @xmath54 along @xmath49 by angle @xmath72 " .
the bending map @xmath249 is continuous and unique up to the post - composition with an element of @xmath9 .
in addition , @xmath249 is isometric on each leaf of @xmath49 and on the closure of each complementary region of @xmath52 .
the map @xmath249 is @xmath4-smooth in the complement of the leaves of @xmath47 with positive weight ( note that the complement has the full lebesgue measure ) . roughly speaking , if @xmath250 are sufficiently close to each other , then the hyperbolic hyperbolic tangent planes of @xmath249 at @xmath60 and @xmath61 intersect at the external angle approximately equal to @xmath251 ( with respect to the normal vector field of @xmath252 ; c.f .
the hyperbolic tangent planes defined in [ thp ] ) .
assume that @xmath23 is an isolated leaf of @xmath49 and that @xmath179 and @xmath253 are two adjacent complementary regions of @xmath52 separated by @xmath23 .
then @xmath254 and @xmath255 are isometric copies of @xmath256 and @xmath257 that intersect at the external angle @xmath78 .
this property determines the bending map @xmath249 if @xmath47 consists of isolated leaves . for general @xmath47
, there is a sequence @xmath258 of measured laminations on @xmath53 consisting of finitely many leaves , that approximates @xmath47 in thurston s topology .
then @xmath259 uniformly on compacts .
( for details , see @xcite . )
let @xmath63 be a simply connected ( open ) region in @xmath8 .
then @xmath63 can be regarded as a projective structure on @xmath260 .
consider @xmath261 , the convex hull of @xmath262 in @xmath54 .
it turns out that @xmath263 is isometric to @xmath53 with respect to the induced path metric on @xmath263 .
there is a unique measured lamination @xmath47 on @xmath53 such that @xmath47 does not contain leaves of weight more than @xmath264 and its bending map @xmath249 realizes the isometry from @xmath53 to @xmath263 .
then , by the orthogonal projection along geodesics in @xmath54 , @xmath265 maps onto @xmath266 .
let @xmath41 be the space of all projective structures on @xmath0 .
william thurston gave a parametrization of @xmath41 that reflects the geometry of projective structures in a combinatorial manner .
this parametrization is useful for the proof of the main theorems of this paper , since it involves a decomposition of @xmath30 into @xmath14-injective regions , where @xmath267 is the developing map of a projective structure .
@xmath41 is naturally homeomorphic to the product of the teichimller space of @xmath0 and the space of measured laminations on @xmath0 : @xmath268 ( the proof is in @xcite . for the following discussion ,
see also @xcite . )
below we discuss some properties of this homeomorphism .
an element in the left hand side of ( [ eq1 ] ) is a pair @xmath11 , where @xmath267 and @xmath31 .
take a pair @xmath269 .
then the total lift @xmath145 of @xmath270 induces a bending map @xmath271 .
furthermore , since the action of @xmath33 preserves @xmath56 , @xmath272 induces a representation @xmath273 such that @xmath252 is @xmath274-equivariant . if @xmath11 and @xmath270 represent the same projective structure , then @xmath275 . letting @xmath276
be the trivial representation , @xmath277 is a projective structure on @xmath55 , which is the universal cover of @xmath1 .
moreover , @xmath277 corresponds to the measured lamination @xmath145 through the orthogonal projection and the bending map , which generalizes the correspondence between a simply connected region in @xmath8 and an injective bending map discussed in
namely , in our current case , @xmath14 and @xmath252 are not necessarily embeddings , and we need to divide the domain of @xmath14 and the domain of @xmath252 so that their corresponding subdomains are homeomorphic through the orthogonal projection .
we shall first discuss the same correspondence for projective structures on @xmath55 , which is more general than the above case .
namely , there is a bijective correspondence between the projective structures on @xmath55 ( that are not conformally equivalent to the euclidian plane ) and the measured laminations on @xmath53 ( up to the action of @xmath278 ) . for a measured lamination @xmath71 on @xmath53 ,
let @xmath279 denote the corresponding projective structure on @xmath55 .
we shall discuss the correspondence between @xmath280 and @xmath47 .
there are a ( topological ) measured lamination @xmath281 on @xmath282 and the _ collapsing map _
@xmath283 , which describe the subdivision and the orthogonal projections . for each leaf @xmath23 of @xmath47 with positive weight
, @xmath284 is a crescent of angle @xmath78 with the canonical foliation ( compare @xcite ) .
conversely , each crescent of angle @xmath285 in @xmath2 projects to a leaf of weight @xmath285 via @xmath286 in the way discussed in [ flatcylinders ] . in the complement of such crescents
, @xmath286 is an isomorphism , i.e. a @xmath4-diffeomorphism that preserves the measured lamination . in summary , @xmath287
is topologically obtained from @xmath47 by blowing up each leaf @xmath23 with positive weight of @xmath47 as above .
( note that there is no periodic leaves with positive weight of @xmath287 . )
the collapsing map @xmath286 is a continuous surjective map that homeomorphically takes each leaf of @xmath287 to a leaf of @xmath47 and each component of @xmath288 to a component of @xmath289 .
furthermore , this correspondence is bijective except the correspondence between the leaves of the crescents and the leaves of positive weight .
a _ maximal ball _ of a projective structure @xmath1 on @xmath55 is a maximal open subset of @xmath55 that @xmath14 homeomorphically takes to a round open disc in @xmath8 , where the maximality is defined with respect to the set inclusion .
if @xmath290 is a maximal ball , then @xmath291 is a round circle in @xmath8 , and @xmath292 is a copy of @xmath53 whose ideal boundary is @xmath291 .
let @xmath293 , and let @xmath294 be the orthogonal projection along geodesics in @xmath54 .
let @xmath253 be the closure of a component of @xmath295 , or a leaf of @xmath287 that does not bound a component of @xmath295 .
then , @xmath253 is contained in a unique maximal ball @xmath290 and @xmath296 is called the _ core _ of @xmath290 .
conversely , each maximal ball @xmath290 contains a unique core .
these cores of maximal balls form a partition of @xmath55 .
let @xmath271 be the bending map induced by @xmath47 .
then we have @xmath297 on each core @xmath298 , which describes the correspondence of @xmath14 and @xmath252 .
define @xmath299 by @xmath300 when @xmath301 .
then we have @xmath302 .
let @xmath303 be the union of leaves of @xmath47 with positive weight .
recall that @xmath252 is @xmath4-smooth except on @xmath303 .
the * hyperbolic tangent plane * of @xmath304 at @xmath60 is @xmath305 when @xmath301 ( see figure [ tangentplane ] ) .
this tangent plane is a support plane of @xmath306 at @xmath307 where @xmath308 is a sufficiently small neighborhood of @xmath309 .
then this hyperbolic tangent plane coincides with the standard hyperbolic tangent plane at each point of @xmath310 , which is the complement of the disjoint foliated crescents .
when @xmath311 moves infinitesimally , the hyperbolic tangent plane of @xmath304 at @xmath60 rotates about @xmath312 in @xmath53 by the amount of the transversal measure @xmath313 , where @xmath314 is a leaf of @xmath287 through @xmath60 if it exists .
in particular , when @xmath60 moves along a leaf or a moves in the closure of a component of @xmath315 , then the hyperbolic tangent plane does not change .
below we show that the hyperbolic tangent planes depend continuously on @xmath316 .
since @xmath252 is @xmath4-smooth away from the isolated leaves of @xmath317 , continuity of the hyperbolic tangent planes on the complement to the union of crescents is clear .
consider a crescent @xmath318 , where @xmath23 is an isolated leaf of @xmath317 .
let @xmath189 ( @xmath319 ) denote the closures of the components of @xmath320 . for each @xmath319 ,
the restriction @xmath321 is @xmath4-smooth along @xmath23 .
therefore , there is a unique hyperbolic plane in @xmath322 through @xmath323 which shares the tangent plane with the surface @xmath324 at the point @xmath323 .
these hyperbolic tangent planes of @xmath324 at @xmath323 intersect along the geodesic @xmath325 at angle @xmath78 . on the other hand , these hyperbolic tangent planes of @xmath324 agree with the hyperbolic tangent planes of @xmath304 at points on their corresponding boundary leaves of @xmath326 . for each @xmath327 ,
the hyperbolic tangent plane of @xmath304 at @xmath60 contains @xmath325 .
when @xmath60 moves in @xmath326 transversally to the foliation , the hyperbolic tangent plane rotates about @xmath325 by the amount of the transversal measure @xmath313 .
therefore , the hyperbolic tangent plane of @xmath304 depends continuously on @xmath327 , hence depends continuously on @xmath311 .
let us return to the correspondence between a measured lamination @xmath47 on @xmath48 and a projective structure @xmath19 on @xmath0 .
since @xmath328 , we have the canonical lamination @xmath329 on @xmath330 and the collapsing map @xmath331 . by its construction , @xmath329 is invariant under the action of @xmath33 , and it induces a measured lamination @xmath332 on @xmath0 .
in addition , @xmath333 is @xmath15-equivariant , and it induces the _ collapsing map _ @xmath334 .
accordingly , for each periodic leaf @xmath23 of @xmath47 , @xmath284 is a flat cylinder of height @xmath78 with the canonical foliation .
conversely , each foliated flat cylinder of height @xmath285 in @xmath2 projects to a periodic leaf of weight @xmath285 via @xmath286 . in the complement of such flat cylinders
, @xmath286 is an isomorphism .
figure [ f&kappa&beta ] illustrates the basic case when we have a measured lamination consisting of a periodic leaf on a complete hyperbolic cylinder . near a periodic leaf of @xmath47 on @xmath48
, we locally have a similar diagram .
let @xmath71 be a measured lamination on @xmath53 .
let @xmath63 be a geodesic or a convex subset of @xmath53 bounded by geodesics .
the * intersection * of @xmath47 and @xmath63 is a measured lamination @xmath335 on @xmath53 , where @xmath336 and @xmath337 is the restriction of @xmath72 to @xmath338 .
denote the intersection by @xmath339 .
let @xmath63 be a convex subset of @xmath53 bounded by geodesics .
the * restriction * of @xmath47 to @xmath63 is a measured lamination @xmath340 on @xmath63 , where @xmath341 and @xmath342 is defined by @xmath343 for all geodesic segments @xmath76 in @xmath63 .
denote the restriction by @xmath344 .
let @xmath345 be a projective structure on a surface @xmath346 , and let @xmath347 be a subsurface of @xmath346 .
then the * restriction * of @xmath345 to @xmath347 is the restriction of the atlas of @xmath345 to @xmath347 .
let @xmath348 denote the restriction of @xmath345 to @xmath347 .
conversely , if a projective structure @xmath349 on a surface @xmath350 is isomorphic to the one obtained by restricting @xmath345 to a subsurface of @xmath346 , then we say that @xmath349 * embeds * into @xmath345 .
let @xmath351 and assume , in addition , that the inclusion @xmath352 is @xmath353-injective .
then the above definition is equivalent to the following : the restriction @xmath348 is the projective structure @xmath354 , where @xmath355 is the restriction of @xmath356 to a lift @xmath357 of @xmath347 to the universal cover of @xmath346 and @xmath358 is the restriction of @xmath359 to @xmath360 acting on @xmath357 .
let @xmath47 be a measured lamination on @xmath53 , and let @xmath361 , the projective structure on @xmath55 corresponding to @xmath47 ( [ thp ] ) .
let @xmath63 be a convex subset of @xmath53 bounded by geodesics , allowing @xmath63 to possibly be a geodesic .
let @xmath362 , and let @xmath363 .
we also let @xmath364 and @xmath365 be the collapsing maps for @xmath280 and @xmath366 , respectively . [ c(i)/stc(l ) ] there exists a homeomorphism @xmath367 such that @xmath368 on @xmath369 .
moreover , @xmath366 embeds into @xmath280 .
consider the leaves of @xmath47 and components of @xmath370 that intersect @xmath63 , and let @xmath371 be the union of these leaves and components .
then @xmath371 is a convex subset of @xmath53 containing @xmath63 , and it is bounded by some leaves of @xmath47 .
we also have @xmath372 .
therefore , it suffices to prove the lemma for @xmath371 .
let @xmath287 and @xmath373 be the canonical measured laminations on @xmath374 and @xmath375 , respectively . since @xmath376
, we can assume that @xmath377 on @xmath371 , where @xmath249 and @xmath378 are the bending maps induced by @xmath47 and @xmath379 , respectively .
therefore , @xmath380 on @xmath381 by the continuity of bending maps .
recall that @xmath382 and @xmath383 are obtained from @xmath371 in the exactly same way , namely by blowing up the periodic leaves of @xmath384 .
therefore , we have a canonical homeomorphism @xmath385 such that @xmath386 isomorphically takes @xmath387 to @xmath388 and @xmath389 on @xmath390 . furthermore , the hyperbolic tangent plane of @xmath391 at @xmath392 coincides with the hyperbolic tangent plane of @xmath393 at @xmath394 .
( see [ thp ] . )
the maximal ball of @xmath395 whose core contains @xmath60 maps to a round open disk by @xmath396 , and the maximal ball of @xmath397 whose core contains @xmath398 maps to a round open disk by @xmath399 .
the convex hull boundaries of these open desks are the hyperbolic tangent planes of @xmath393 at @xmath398 and of @xmath391 at @xmath60 , and therefore , they must agree .
then the round disks on @xmath8 also coincide . recall that @xmath400 and @xmath401 are connected by a geodesic in @xmath54 orthogonal to the hyperbolic tangent plane , and so are @xmath402 and @xmath403 . hence , @xmath404 .
each component @xmath405 of @xmath406 is an open or closed half plane bounded by a leaf @xmath23 of @xmath47 , and it does not contain leaves of @xmath379 . then @xmath407 is a component of @xmath408 .
this component is simply connected and bounded by a leaf @xmath409 of @xmath287 that maps to @xmath23 via @xmath410 .
similarly , @xmath411 is a component of @xmath412 .
this component is simply connected and bounded by a boundary curve @xmath413 of @xmath382 .
in addition , @xmath413 is a leaf of @xmath373 or contained in a component of @xmath414 , and it is homeomorphic to @xmath409 and @xmath23 via @xmath386 and @xmath415 , respectively .
the leaf @xmath409 is contained in a unique maximal ball @xmath290 of @xmath280 , whose convex hull boundary is the hyperbolic tangent plane of @xmath393 at each point in @xmath409 .
similarly , @xmath413 is contained in a unique maximal ball @xmath416 of @xmath366 , whose convex hull boundary is the hyperbolic tangent plane of @xmath391 at each point in @xmath413 .
thus these hyperbolic tangent planes are the same planes in @xmath54 .
therefore , we can identify @xmath416 and @xmath290 by a @xmath4-diffeomorphism @xmath417 such that @xmath418 on @xmath416 .
then @xmath419 on @xmath420 .
therefore , the embedding @xmath421 extends to @xmath411 , preserving @xmath368 .
the different components of @xmath412 map to different components of @xmath408 by the extension . hence , since @xmath422 , we have an embedding of @xmath366 into @xmath280 .
assume that @xmath63 has non - empty interior and no boundary geodesic of @xmath63 transversally intersects a leaf of @xmath47 with positive weight .
then @xmath423 and @xmath424 are @xmath4-smooth subsurfaces of @xmath55 .
thus we immediately obtain [ basiccoro ] @xmath425 and @xmath426 are isomorphic as projective structures .
let @xmath71 be an irrational minimal lamination on a closed orientable hyperbolic surface @xmath0 , and @xmath50 be the sequence of simple loops on @xmath0 that converges to @xmath52 constructed in [ appro ] .
let @xmath427 be the total lift of @xmath47 to @xmath53 , and let @xmath57 be a lift of @xmath51 to @xmath53 .
let @xmath428 be @xmath429 ( see [ intersection ] ) .
note that the dual tree of @xmath430 is isometric to @xmath148 .
let @xmath87 be a measured lamination on a hyperbolic surface @xmath3 .
define the * norm * of @xmath89 by @xmath431 where @xmath76 varies over all geodesic segments of length less than @xmath432 on @xmath3 .
we next prove that the transversal measure @xmath433 of short geodesic segments is bounded by an arbitrary small number , provided that @xmath37 is large : [ propa ] @xmath434 .
the basic idea of the proof is that , when a geodesic segment @xmath76 with @xmath435 intersects a measured lamination at an angle close to zero , its transversal measure is also close to zero .
let @xmath60 be a point on a leaf @xmath23 of @xmath143 . for @xmath436 $ ] ,
let @xmath437 be the geodesic on @xmath30 intersecting @xmath23 at @xmath60 with @xmath438 ( see [ appro ] ) .
set @xmath439 .
[ ml2 ] for every @xmath440 , there exists @xmath441 such that , if @xmath442 and @xmath443 , then @xmath444 for all geodesic segments @xmath76 in @xmath53 with @xmath445 and @xmath435 . for an arbitrary @xmath61 in @xmath155 ,
let @xmath23 be the leaf of @xmath143 through @xmath61 . then consider @xmath446 , the closed hyperbolic disk of radius @xmath29 centered at @xmath61 . in each component of @xmath188
, choose a geodesic @xmath447 close to @xmath23 that does _ not _ transversally intersect a leaf of @xmath143 , i.e. @xmath448 is a leaf of @xmath143 or is in the complement of @xmath155 .
let @xmath449 be the convex region bounded by @xmath450 and @xmath451 , which contains @xmath23 .
for every @xmath181 , by applying lemma [ box - nhd ] to @xmath452 , we can assume that @xmath450 and @xmath451 are close enough to @xmath23 , so that the @xmath453 is contained in a flow box of height less than @xmath182 .
take a neighborhood @xmath290 of @xmath61 whose closure is contained in the interior of @xmath454
. then there exists ( small ) @xmath441 such that , if @xmath455 and @xmath456 , then @xmath457 ( see figure [ r - u - theta ] ) .
since @xmath453 is contained in the flow box of height less than @xmath182 and @xmath253 supports @xmath458 , for every geodesic segment @xmath76 in @xmath446 , the transversal measure of @xmath76 with respect to @xmath458 is bounded by @xmath182 .
therefore , since @xmath459 is a sublamination of @xmath460 , @xmath461 .
if @xmath76 is a geodesic segment in @xmath53 such that @xmath462 and @xmath435 , then @xmath463 . thus , @xmath461
this proves the lemma if @xmath60 is in @xmath290 , which is a neighborhood of @xmath61 .
since @xmath0 is compact and @xmath56 is invariant under the deck transformations , the lemma follows .
( proposition [ propa ] ) fix arbitrary @xmath181 .
it suffices to show that , for sufficiently large @xmath37 , if a geodesic segment @xmath76 in @xmath53 satisfies @xmath464 , then @xmath465 . by lemma [ ml2 ]
, there exists @xmath441 , such that , if @xmath466 and @xmath467 , then @xmath468 for every geodesic segment @xmath76 with @xmath445 and @xmath435 . by lemma [ ml1 ] , for sufficiently large @xmath37 , @xmath469 and @xmath470 .
consider the cyclic subgroup of @xmath471 generated by the translation along @xmath57 by @xmath236 , which we can regard as @xmath472 acting on @xmath53 .
choose a leaf @xmath473 , and consider the orbit of @xmath195 under the action of the cyclic group .
this orbit forms a sublamination @xmath90 of @xmath474 .
the leaves of @xmath90 intersect @xmath57 at a constant angle less than @xmath475 , and the distance between their consecutive intersection points equals to @xmath236 . since @xmath470 and
the dual tree of @xmath430 is isometric to @xmath148 , if a geodesic segment @xmath76 lies between two adjacent leaves of @xmath90 , then @xmath476 .
therefore , if @xmath464 , then @xmath76 transversally intersects @xmath91 at least three times .
let @xmath477 be the intersection points lying on @xmath76 in this order , and let @xmath478 be the leaves of @xmath90 through these points . take @xmath479 such that @xmath480 is the maximal odd integer _ not _ exceeding @xmath147 .
let @xmath68 be the subsegment of @xmath76 with end points @xmath481 and @xmath482 .
( see figure [ s&r ] . )
then @xmath483 and @xmath484 .
let @xmath485 be the geodesic segment that realizes the distance between @xmath486 and @xmath487 .
then @xmath485 is orthogonal to @xmath486 and @xmath487 .
in addition , @xmath485 intersects @xmath58 transversally , since otherwise @xmath488 bound a hyperbolic triangle whose interior angle sum is @xmath264 or a hyperbolic rectangle whose interior angle sum is @xmath489 .
therefore , the triangle bounded by @xmath58 , @xmath485 , @xmath486 is isometric to the triangle bounded by @xmath490 .
thus @xmath491 is the middle point of @xmath485 .
note that @xmath492 , @xmath493 intersects @xmath494 at @xmath491 at an angle less than @xmath475 , and @xmath495 .
therefore , @xmath496 .
hence , @xmath497 .
in this section , let @xmath71 be a measured lamination on @xmath53 and @xmath248 be the bending map induced by @xmath47 . recall that @xmath498 , where @xmath76 varies over all geodesic segments on @xmath53 of length less than 1 .
[ ib ] there exists @xmath499 such that , if @xmath500 , then the induced bending map @xmath249 is a bilipschitz embedding ; hence , it continuously extends to @xmath501 as an embedding whose image is a simple loop in @xmath502 . to prove the decomposition theorem ( theorem [ i d ] )
, we need a generalization of theorem [ ib ] . however , if we do not require flat cylinders in theorem [ i d ] to be integral , then theorem [ ib ] is sufficient . to state the generalization ,
let us set up some notions .
a leaf @xmath23 of @xmath47 is _ outermost _ , if the other leaves of @xmath47 lie only in one component of @xmath188 .
consider all outermost leaves of @xmath47 with positive weight . then this set forms a sublamination @xmath503 of @xmath49 , and @xmath503 consists of isolated leaves .
let @xmath504 be the measured lamination on @xmath53 obtained by assigning each leaf of @xmath503 its positive weight with respect to @xmath72 .
let @xmath281 be a sublamination of @xmath504 .
we also let @xmath505 .
let @xmath506 .
note that @xmath507 bounds a convex region of @xmath53 and that the convex region contains @xmath508 .
[ ib2 ] for every @xmath509 , there exist @xmath510 and @xmath499 with the following property : if a measured lamination @xmath47 on @xmath53 contains a sublamination @xmath287 such that + ( i ) @xmath511 , + ( ii ) @xmath512 , + ( iii ) @xmath513 , and + ( iv ) every leaf of @xmath287 has weight less than @xmath514 , + then the induced bending map @xmath249 is an injective ( s , t)-quasiisometric embedding .
( compare with the proof of ( * ? ? ?
* theorem 4.2.2 ) . ) for arbitrary @xmath509 , choose @xmath515 .
next , choose @xmath516 small enough so that @xmath517 , @xmath518 and @xmath519 . then @xmath520 . let @xmath521 \to { \mathbb{h}}^2 $ ] be an arbitrary geodesic segment parametrized by arc length , where @xmath201 is the length of @xmath147 .
let @xmath522 \to { \mathbb{h}}^3 $ ] . * discrete case . *
first , assume that @xmath47 consists of isolated leaves .
then @xmath147 transversally intersects @xmath80 only finitely many times .
let @xmath523 be the points in @xmath524 that correspond to such transversal intersection points .
the curve @xmath21 is a piecewise geodesic , and @xmath525 are its singular points . in addition
, let @xmath526 and @xmath527 . for @xmath528 , let @xmath51 denote the leaf of @xmath47 intersecting @xmath147 at @xmath529 . for every pair of consecutive geodesic segments @xmath530)$ ] and @xmath531)$ ] , their exterior angle , @xmath532 , is bounded from above by @xmath533 , by the triangle inequality .
since @xmath507 bounds a convex region that contains @xmath508 , @xmath534 transversally intersects @xmath507 at most twice at @xmath535 and/or @xmath536 .
define @xmath537 \to { \mathbb}{r}_{\geq 0}$ ] by @xmath538 .
define @xmath539 \to [ 0,\pi]$ ] by @xmath540 and @xmath541 when @xmath542 satisfies @xmath543 ( figure [ deftheta ] ) .
call @xmath544 the _ angle function _ for @xmath21 .
then @xmath545 is smooth in @xmath546 . at @xmath525 with @xmath547 ,
@xmath548 is smooth from the left and discontinuous from the right .
let @xmath549 .
then , by the bound for the exterior angle and the triangle inequality , @xmath550 we also have @xmath551 ( see ( * ? ? ?
* i.4.2.12 . ) , @xcite ) .
in particular , @xmath545 does not increase on @xmath552 $ ] .
note that @xmath553 and @xmath554 completely determine the change in @xmath98 .
therefore , by ( [ jump ] ) , the total increase of @xmath98 from its discontinuity points in @xmath555 \cap [ 0 , p]$ ] is at most @xmath556 for all @xmath542 .
let @xmath557\ > |\ > \theta(t ) > \theta_0 + \delta \}$ ] , and let @xmath558 denote the lebesgue measure of @xmath559 .
the following proposition yields good upper bounds for @xmath558 and @xmath544 .
[ l1 ] there exist @xmath560 and @xmath561s , which depend only on @xmath562 and @xmath563 , such that @xmath564 and @xmath565 for all geodesic segments @xmath521 \to { \mathbb{h}}^2 $ ] .
_ remark : _
@xmath566 and @xmath1 do not depend on @xmath47 and @xmath287 as long as they satisfy the hypotheses ( i ) - ( iv ) of theorem [ ib2 ] .
the following two lemmas are used in the proof of the proposition .
let @xmath567 be a subinterval of @xmath568 $ ] in which @xmath147 does _ not _ transversally intersect leaves of @xmath287 .
then @xmath512 bounds the total increase of @xmath98 from the discontinuity points in @xmath555 \cap [ a , b]$ ] for all @xmath569 $ ] .
[ l2 ] if @xmath570 , then @xmath571 for all @xmath572 $ ] .
( this proof is essentially the same as the proof of lemma 4.4 in @xcite . )
assume that @xmath98 is _ not _ bounded by @xmath573 .
then let @xmath574\ > |\ > \theta(t ) \geq \theta_0 + \delta \}$ ] , and let @xmath575 \>|\ > \theta(t ) < \theta_0\}$ ] .
note that @xmath576 are both discontinuity points .
thus @xmath577 .
then @xmath578 for all sufficiently small @xmath181 .
therefore , @xmath579 . if @xmath580 , then @xmath581 . if , in addition , @xmath542 is a smooth point , @xmath582 .
therefore , @xmath583 this inequality contradicts the definition of @xmath584 .
[ l3 ] if @xmath585 , then @xmath586 for all @xmath569 $ ] . _
remark : _ the right hand side of ( [ ib1 ] ) decreases as @xmath542 increases until it reaches @xmath573 . as with lemma [ l2 ]
, we can show that @xmath587 for all @xmath588 $ ] .
if @xmath589 for some @xmath588 $ ] , then let @xmath590 \,|\ , \theta(t ) \leq \theta_0\}$ ] .
by lemma [ l2 ] , @xmath591 for all @xmath592 $ ] . for all @xmath593 , @xmath594 since @xmath595
therefore , @xmath596 if @xmath597 for all @xmath569 $ ] , similarly ( [ en2 ] ) ( [ en3 ] ) hold .
this completes the proof .
( proposition [ l1 ] ) recall that the geodesic segment @xmath147 transversally intersects @xmath507 at most twice at @xmath535 and/or @xmath536 .
thus we have four cases : + _ case 1 .
_ @xmath598 does not transversally intersect @xmath507 .
+ _ case 2 .
_ @xmath598 transversally intersects @xmath507 only at @xmath535 .
+ _ case 3 .
_ @xmath598 transversally intersects @xmath507 only at @xmath536 . + _ case 4 .
_ @xmath598 transversally intersects @xmath507 at both @xmath535 and @xmath536 .
+ in every case , @xmath599 in @xmath600 $ ] , and @xmath601 = \emptyset$ ] . _ case 1 . _
since @xmath602 and @xmath147 does not transversally intersects @xmath507 , by applying lemma [ l2 ] with @xmath603 and @xmath604 , we have @xmath605 for all @xmath606 $ ]
. therefore , @xmath607 .
hence , we can take arbitrary @xmath608 and @xmath609 .
_ since @xmath610 , @xmath611 .
in @xmath612 $ ] , by taking the limit of ( [ ib1 ] ) as @xmath613 , we obtain @xmath614 then , if @xmath615 , @xmath616 therefore , @xmath617 ) = mes(i_p)$ ] is bounded from above by the right hand side of ( [ en7 ] ) .
we thus can take @xmath618 _ case 3 .
_ on the interval @xmath619 $ ] , by lemma [ l2 ] , @xmath620 . since @xmath621 , @xmath622 . on the interval
@xmath623 $ ] , @xmath548 does not increase , and we obtain @xmath624 .
therefore , @xmath98 is bounded by @xmath625 for all @xmath606 $ ] . if @xmath615 , then @xmath626 and , since @xmath627 is convex on @xmath628 $ ] , @xmath629 then @xmath630 therefore , if @xmath631 , then @xmath632 thus @xmath633 ) = mes(i_p)$ ] is bounded by the right hand side of ( [ en6 ] ) .
_ this case is basically the combination of case 2 and case 3 . in @xmath634
$ ] , by an argument similar to that of case 2 , @xmath548 is bounded by @xmath635 .
in addition , the inequalities ( [ en4.5 ] ) , ( [ en5 ] ) hold , and @xmath636)$ ] is bounded from above by the right hand side of ( [ en7 ] ) .
therefore , if @xmath637 , @xmath638 by hypothesis ( iii ) and the relation between @xmath639 and @xmath640 , @xmath641 let @xmath642 .
then @xmath643 . on @xmath623
$ ] , since @xmath98 does not increase , @xmath644 .
if @xmath615 , then @xmath645 therefore , @xmath633)$ ] is bounded by the right hand side of ( [ en8 ] ) .
hence , @xmath558 is bounded by the sum of the right hand sides of @xmath646 and @xmath647 , and @xmath98 is bounded by @xmath648 .
in each case 1 case 4 , we have found upper bounds of @xmath544 and @xmath558 that do not depend on the choices for @xmath147 , @xmath47 and @xmath287 . the proposition immediately follows . _
( continue the proof of theorem [ ib2 ] . )
_ let @xmath649 and @xmath650 , which are the end points of the geodesic segment @xmath521 \to { \mathbb{h}}^2 $ ] .
then @xmath651 .
we have @xmath652 { \setminus}i_p } \cos(\theta(t ) ) dt + \int_{i_p } \cos(\theta(t))dt.\end{aligned}\ ] ] when @xmath653 , @xmath654 , while by proposition [ l1 ] , @xmath558 is bounded by @xmath1 .
therefore , @xmath655 on the other hand , @xmath656 recall that @xmath147 is an arbitrary geodesic in @xmath53 .
accordingly , @xmath60 and @xmath61 are arbitrary points in @xmath53 .
therefore , letting @xmath657 we conclude that @xmath252 is an @xmath658-quasiisometric embedding .
we claim that @xmath252 is injective .
if @xmath659 for distinct @xmath660 , let @xmath147 be the geodesic segment connecting @xmath60 to @xmath61 .
then @xmath661}$ ] is the geodesic segment connecting @xmath662 to @xmath60 .
then , for @xmath663 , @xmath664 .
this contracts the upper bound @xmath665 of @xmath98 .
therefore , @xmath252 is injective .
* general case .
* for the rest of the proof , we deal with the case that @xmath47 does _ not _ consist of isolated leaves . take a sequence of measured laminations on @xmath53 , @xmath666 , such that each @xmath430 consists of finitely many leaves and @xmath258 limits to @xmath47 in thurston s topology .
let @xmath667 denote the bending map induced by @xmath430 .
then @xmath668 converges to @xmath249 uniformly on compacts .
( see @xcite for the construction of @xmath258 and the convergence of @xmath669 . ) for an arbitrary geodesic segment @xmath521 \to { \mathbb{h}}^2 $ ] , let @xmath670 \to { \mathbb{h}}^3 $ ] . by the discrete case
, we have @xmath671 for sufficiently large @xmath37 .
in addition , by the convergence of @xmath668 to @xmath249 , @xmath672 limits to @xmath673 as @xmath674 . then @xmath675 therefore , @xmath249 is an @xmath676-quasiisometric embedding .
we claim that @xmath249 is injective .
suppose that @xmath249 is not injective .
then there exist distinct @xmath660 with @xmath677 .
let @xmath521 \to { \mathbb{h}}^2 $ ] be the geodesic segment from @xmath60 to @xmath61 .
there are only countably many leaves of @xmath47 with positive weight ( in particular , near @xmath61 ) .
note that the transversal measure restricted to the other leaves of @xmath47 is non - atomic .
therefore , @xmath678 limits to @xmath124 as @xmath679 ( recall that , if @xmath680 is contained in a leaf of @xmath47 with positive weight , then the weight does not contribute to the transversal measure @xmath681 by our convention ) . thus , for every small @xmath181 , there exists @xmath682 such that @xmath683 .
therefore , for every @xmath181 , there exist @xmath684 such that , letting @xmath685 @xmath686 , @xmath687 , @xmath688 and @xmath689 for sufficiently large @xmath37 , @xmath690 by the convergence of @xmath433 to @xmath72 .
recall that @xmath691 , where @xmath668 is the bending map induced by @xmath430 .
let @xmath692 , @xmath693 $ ] be the non - smooth points of @xmath235 closest to @xmath694 in @xmath695 and @xmath696 $ ] , respectively .
the convergence of @xmath258 implies that , for sufficiently large @xmath37 , @xmath697 bounds from above the sum of the external angles at the non - smooth points of @xmath235 on @xmath698 . in the following
, we always assume that @xmath182 is sufficiently small . since @xmath235 is almost straight " on @xmath699 , for every @xmath700 , @xmath701 provided that @xmath37 is sufficiently large .
( see figure [ beta_l ] . ) for sufficiently large @xmath37 , we also have @xmath702 by ( [ xz_1<<z_1z_2 ] ) , and @xmath703 since @xmath704 and @xmath705 are close to each other and far from @xmath706 .
if @xmath235 is not smooth at @xmath694 , the external angle of @xmath235 at @xmath694 is @xmath707 , and , since @xmath708 , the external angle is bounded from above by @xmath182 .
let @xmath709 be the angle function of @xmath235 .
then , by the triangle inequality , @xmath710 ( see figure [ beta_l ] ) . therefore , by taking sufficiently small @xmath182 and @xmath711 , @xmath712 for large @xmath37 , where @xmath561 is the constant obtained from proposition [ l1 ] .
contradictorily , by proposition [ l1 ] , we can show @xmath713 for large @xmath37 .
assume that @xmath47 satisfies the assumptions of theorem [ ib ] or theorem [ ib2 ] . by these theorems
, @xmath714 is an injective quasiisometric embedding , and , hence , it extends continuously to a homeomorphism @xmath715 from @xmath716 onto a simple loop on @xmath717 ( see @xcite , @xcite ) .
therefore , @xmath718 consists of two simply connected regions .
[ admissibled ] under the assumption of theorem [ ib ] or theorem [ ib2 ] , the projective structure @xmath280 on @xmath719 corresponding to @xmath47 is admissible . since @xmath252 is an injective continuous quasiisometric embedding
, @xmath720 is a proper surface embedded in @xmath54 .
therefore , @xmath720 separates @xmath54 into two components ( the jordan - brouwer separation theorem ) .
since @xmath720 is locally convex , one of the components of @xmath721 is convex ( see ( * ? ? ?
* i.1.3 ) ) .
the concave component of @xmath721 is cobounded by @xmath720 and a topological closed disk @xmath640 contained in @xmath8 . then @xmath722 .
since @xmath47 does not contain leaves with weight @xmath723 , @xmath724 is the canonical bending lamination on @xmath725 .
therefore , @xmath200 is thurston s coordinates for the projective structure on @xmath726 ; hence @xmath280 is admissible .
let @xmath0 be a closed orientable surface of genus at least @xmath29 .
let @xmath19 be a projective structure on @xmath0 .
express @xmath1 in thurston s coordinates as @xmath270 , where @xmath46 is a marked hyperbolic structure on @xmath0 and @xmath71 is a measured lamination on @xmath48 .
let us recall other related notions from [ p ] : let @xmath427 be the total lift of @xmath47 to @xmath53 , and let @xmath727 be the bending map induced by @xmath56 .
let @xmath728 be the collapsing map , and let @xmath729 be the lift of @xmath730 to a map between the universal covers of @xmath2 and @xmath48 .
let @xmath281 be the canonical ( topological ) measured lamination on @xmath2 corresponding to @xmath47 via @xmath286 ( [ thp ] ) .
let @xmath329 be the total lift of @xmath287 to @xmath330 .
since @xmath47 decomposes into minimal measured laminations , we can set @xmath731 where @xmath732 are the periodic minimal sublaminations of @xmath47 and @xmath733 are the irrational minimal sublaminations of @xmath47 .
let @xmath734 denote the periodic leaf supporting @xmath735 , and let @xmath147 denote the periodic part of @xmath52 , @xmath736 let @xmath87 denote the irrational part of @xmath47 , @xmath737 for each @xmath738 , let @xmath739 be the sequence of simple geodesic loops on @xmath740 that approximates @xmath741 , constructed in [ appro ] .
by lemma [ ml1 ] , we have @xmath742 .
since @xmath741 is an isolated subset of @xmath52 , we can assume that @xmath743 does not intersect @xmath744 .
therefore , @xmath745 . let @xmath746 , so that @xmath747 . recall that , for each @xmath285 , @xmath748 is a flat foliated cylinder of height @xmath749 in @xmath750 , where @xmath749 is the weight of @xmath751 .
the foliation on @xmath752 consists of admissible loops that are homeomorphic to @xmath751 via @xmath286 ( [ thp ] ) .
recall also that @xmath286 is a @xmath4-diffeomorphism on @xmath753 .
therefore , if @xmath23 is an essential simple loop on @xmath48 disjoint from @xmath52 , then @xmath754 is an essential simple loop on @xmath2 disjoint from @xmath755 .
we shall see that @xmath754 is also admissible .
let @xmath201 be the component of @xmath756 containing @xmath23 , and let @xmath24 and @xmath757 be the lifts of @xmath23 and @xmath201 , respectively .
we can assume that @xmath758 and that @xmath759 acts on @xmath24 freely and properly discontinuously as an infinite cyclic subgroup of @xmath9 generated by a hyperbolic element .
let @xmath26 be the homotopy class of @xmath23 that generates @xmath759 . since @xmath24 is a quasigeodesic in @xmath53 and @xmath760 is an isometry on @xmath757
, @xmath761 is a quasigeodesic in @xmath54 .
then @xmath759 acts on @xmath762 freely and properly discontinuously via @xmath15 , and this action fixes the distinct end points of @xmath762 on @xmath8 .
therefore , @xmath25 is a loxodromic element of @xmath9 .
the curve @xmath763 is a lift of @xmath764 to @xmath330 .
then @xmath763 is contained in @xmath765 , which is a component of @xmath766 . since @xmath767 is the core of a maximal ball , @xmath763
is contained in a maximal ball .
thus , @xmath14 is an embedding on @xmath763 .
hence , @xmath754 is admissible .
we have @xmath768 and @xmath769 . for each @xmath770 , let @xmath771 , which is a simple loop on @xmath2 . the fact that @xmath745 suggests the following proposition : for each @xmath772 ,
@xmath773 is admissible , provided that @xmath37 is sufficiently large .
let @xmath774 be a lift of @xmath743 to @xmath53 .
consider the measured lamination @xmath775 . since @xmath743 is disjoint from @xmath744 , @xmath776 , where @xmath777 is the total lift of @xmath778
choose @xmath516 as in theorem [ ib ] .
applying proposition [ propa ] with @xmath779 , we have @xmath780 for all large @xmath37 . by theorem [ ib ] , for sufficiently large @xmath37
, the bending map @xmath781 induced by @xmath782 is an injective quasiisometric embedding , and it continuously extends to an embedding of @xmath716 .
let @xmath783 be the homotopy class of @xmath743 that acts on @xmath53 as a hyperbolic element of @xmath278 preserving @xmath774 .
the extension of @xmath784 homeomorphically takes the limit set of @xmath785 to the limit set of @xmath786 .
thus @xmath787 is loxodromic . by corollary
[ admissibled ] , the projective structure @xmath788 on @xmath55 corresponding to @xmath775 is admissible for sufficiently large @xmath37 .
then the developing map @xmath399 of @xmath788 is an embedding . by lemma [ c(i)/stc(l ) ]
, there exists a homeomorphism @xmath789 such that @xmath790 on @xmath791 , where @xmath792 is the collapsing map for @xmath788 .
since @xmath399 is an embedding , @xmath793 restricted to @xmath774 is an embedding .
we thus obtain an admissible loop from every minimal sublamination of @xmath47 and every complementary region of @xmath52 that is not topologically an open disk .
therefore , [ gc ] for every projective structure @xmath1 on @xmath0 , there exists an admissible loop on @xmath2 . _
equivalently , we can state that every projective structure on @xmath0 admits a grafting operation .
we carry over our notation from the previous section . we have shown that @xmath794 and @xmath751 correspond to admissible loops on @xmath0 through @xmath286 , provided that @xmath37 is sufficiently large .
their union @xmath795 is a multi - loop on @xmath48 . in this section ,
we show that @xmath796 decomposes @xmath2 into admissible subsurfaces .
[ i d ] let @xmath1 be a projective structure on a closed orientable surface @xmath0 of genus at least 2 .
then there exists a decomposition of @xmath0 into cylinders and compact subsurfaces of negative euler characteristic , such that the restriction of @xmath1 to each cylinder is an integral flat structure and the restriction to each subsurface of negative euler characteristic is an admissible projective structure .
note that every flat cylinder of height less than @xmath116 is admissible .
therefore , every integral flat cylinder can be further decomposed into admissible flat cylinders , if we wish .
moreover , by further decomposing each surface of negative euler characteristic into pairs of pants , if necessary , we immediately obtain : [ pants ] there exists a decomposition of @xmath0 into pairs of pants and cylinders such that the restriction of @xmath1 to each cylinder is an integral flat structure and the restriction to each pair of pants is an admissible structure . let @xmath23 be a geodesic lamination on a complete hyperbolic surface @xmath3 .
let @xmath797 denote the collection of all geodesic segments of length less than one on @xmath3 that do _ not _ transversally intersect any leaves of @xmath23 .
then , a geodesic segment @xmath76 connecting @xmath60 and @xmath61 on @xmath3 is an element of @xmath797 if and only if either @xmath798 or @xmath799 .
we claim that , for every @xmath804 , there exist a neighborhood @xmath805 of @xmath60 and @xmath806 such that , if @xmath807 , then @xmath802 for every @xmath808 with @xmath809 .
this would imply the lemma , since @xmath0 is compact .
let @xmath810 and @xmath57 denote the total lifts of @xmath89 and @xmath51 to @xmath53 , respectively .
choose a lift @xmath811 of @xmath60 to @xmath53 . through the covering map from @xmath53 to @xmath48 ,
the above claim is equivalent to the following : there exist a neighborhood @xmath812 of @xmath811 and @xmath813 such that , if @xmath814 , then @xmath815 for every @xmath816 such that @xmath817 .
_ suppose first that @xmath818 ; then , @xmath819 . let @xmath201 be the component of @xmath820 that contains @xmath811
. then @xmath201 is an open convex region bounded by some leaves of @xmath821 .
clearly , only finitely many such boundary leaves intersect @xmath822 .
let @xmath823 denote these intersecting leaves .
let @xmath824 for all @xmath825 .
( see figure [ pandv ] . )
recall that there is a projection @xmath157 , where @xmath156 is the tree dual to @xmath826 ( [ dt ] ) . by lemma [ box - nhd ] , for every @xmath181 and every @xmath827
, there exists a flow box neighborhood @xmath828 of @xmath829 that projects to a geodesic segment of length less than @xmath830 in @xmath156 .
let @xmath831 denote the geodesic segment @xmath832 .
since @xmath833 is a boundary geodesic of @xmath201 , the leaves of @xmath826 contained in @xmath201 do not accumulate to @xmath833 .
therefore , by the construction of @xmath828 , we can assume that @xmath834 is an end point of @xmath831 .
therefore , @xmath835 projects to @xmath836 , the one point union of @xmath831 that identifies the end points @xmath837 of @xmath831
. then @xmath838 also projects onto @xmath839 , and the diameter of @xmath839 is less than @xmath182 .
since the sequence @xmath840 approximates @xmath821 , each @xmath833 is approximated by a sequence @xmath841 such that @xmath842 is a leaf of @xmath57 .
for the rest of case 0 , we always assume that @xmath37 is sufficiently large . for each @xmath37 ,
let @xmath38 be the open convex region bounded by @xmath843 .
then @xmath844 limits to @xmath201 as @xmath37 goes to infinity .
hence , @xmath845 .
take an open neighborhood @xmath812 of @xmath811 such that the closure of @xmath812 is contained in the interior of @xmath846 .
then @xmath812 is contained in @xmath38 . since @xmath847 , every @xmath848 with @xmath817
is contained in @xmath849 , and therefore @xmath76 is contained in @xmath850 . hence @xmath851 and @xmath815 .
next , suppose @xmath852 ; then , @xmath853 . let @xmath854 be the leaf containing @xmath811 , and let @xmath855 .
there are two ways that @xmath821 can accumulate to @xmath23 : + _ case 1 . _
the leaves of @xmath821 accumulate from only one side of @xmath23 , or + _ case 2 . _
the leaves of @xmath821 accumulate from both sides of @xmath23 .
+ in both cases , by lemma [ box - nhd ] , there exists a flow box neighborhood @xmath416 of @xmath856 with height less than @xmath830 . in case 1 , @xmath857 is a boundary geodesic of a component @xmath858 of @xmath820 .
let @xmath859 be the boundary geodesics of @xmath858 that intersect @xmath860 .
we can assume that @xmath861 .
then @xmath23 bounds a half plane disjoint from @xmath858 , and the leaves of @xmath821 contained in this half plane accumulate to @xmath23 .
consider a leaf @xmath195 of @xmath821 in this half plane , such that @xmath862 .
let @xmath201 be the open convex region in @xmath53 bounded by @xmath195 and @xmath863 , which contains @xmath811 .
then @xmath864 is a geodesic segment of length less than @xmath830 .
take an open neighborhood @xmath812 of @xmath811 such that the closure of @xmath812 is contained in the interior of @xmath846 .
the same argument as case 0 shows that @xmath812 satisfies the desired property .
( see the left picture in figure [ case1&2 ] . ) in case 2 , in each component of @xmath188 , consider a leaf @xmath220 , @xmath865 , of @xmath821 close to @xmath23 .
let @xmath201 be the open convex region in @xmath53 bounded by @xmath866 and @xmath867 , which contains @xmath23 . by taking @xmath866 and
@xmath867 sufficiently close to @xmath23 , we can assume that @xmath868 is contained in @xmath416 .
take an open neighborhood @xmath812 of @xmath811 such that the closure of @xmath812 is contained in the interior of @xmath846 .
again , as in case 0 , we see that @xmath812 satisfies the desired property .
( see the right picture in figure [ case1&2 ] . )
let @xmath179 be a component of @xmath869 , and let @xmath870 be a lift of @xmath179 to @xmath53 .
note that , if we take a different lift @xmath870 , then @xmath871 changes only by an element of @xmath33 .
in particular , @xmath872 does not depend on the choice of @xmath870 .
for every @xmath179 , its lift @xmath870 is an open convex region bounded by some leaves of @xmath57 .
therefore , @xmath874 consists of closed half planes bounded by these leaves .
we can assume that @xmath870 is @xmath875-invariant .
let @xmath876 be the number of the boundary components of @xmath179 .
the @xmath875-action permutes the complementary half planes , and this action on the half planes has exactly @xmath876 orbits .
let @xmath877 be the representatives of the orbits .
since @xmath23 consists of @xmath208 disjoint simple loops on @xmath0 , @xmath878 .
let @xmath405 be a component of @xmath874 .
note that @xmath879 is transversal to @xmath826 .
then a leaf @xmath23 of @xmath871 intersects @xmath405 if and only if @xmath23 intersects @xmath879 .
therefore , @xmath880 . by the covering map from @xmath53 to @xmath48
, @xmath881 covers some @xmath743 . recalling that @xmath774 is a lift of @xmath743 to @xmath53 , @xmath882 .
since @xmath871 and @xmath883 coincide on @xmath405 , @xmath884 since @xmath870 is convex , for every geodesic segment @xmath76 in @xmath53 , @xmath885 is either empty or a geodesic segment .
in addition , @xmath886 consists of at most 2 geodesic segments , each of which is contained in a component of @xmath887 ( see figure [ prop7 - 4 ] ) .
therefore , by the definition of the norm , @xmath888 by lemma [ lemid ] , for every @xmath181 , there exists @xmath800 such that , if @xmath801 , then @xmath889 for every component @xmath179 of @xmath869 . by proposition [ propa ] , for every @xmath890 , @xmath891 as @xmath674 . therefore , for every @xmath181 ,
if @xmath37 is sufficiently large , then ( [ eq ] ) is bounded from above by @xmath182 for every @xmath179 .
( theorem [ i d ] ) recall that @xmath195 is the number of the periodic leaves of @xmath47 . for each @xmath892 , let @xmath893 , which is a flat cylinder of height @xmath894 , and let @xmath895 . since @xmath896 is homeomorphic to @xmath897 by @xmath286 , @xmath898 is a simple closed loop on @xmath2 for each sufficiently large @xmath899 and for each @xmath900
. let @xmath901 and @xmath902 .
then @xmath286 restricts to a homeomorphism from @xmath903 to @xmath904 .
in particular , the components of @xmath903 and @xmath904 correspond bijectively .
note that , since @xmath751 s and @xmath743 s are not pairwise homotopic ( when @xmath37 is fixed ) , each component of @xmath903 and @xmath904 is ( the interior of ) a compact surface with boundary of negative euler characteristic .
let @xmath905 .
first , we shall show that , when @xmath37 is large , every component @xmath906 of @xmath907 is admissible .
let @xmath908 , which is a component of @xmath909 .
let @xmath910 and @xmath911 be the corresponding lifts of @xmath253 and @xmath906 to the universal covers , @xmath330 and @xmath145 , respectively , so that @xmath912 .
choose @xmath516 that satisfies theorem [ ib ] .
let @xmath913 . then , since @xmath914 , @xmath915 .
note that @xmath253 is contained in a component of @xmath869 .
therefore , by proposition [ propid ] , @xmath916 for every component @xmath906 of @xmath917 , provided that @xmath37 is sufficiently large .
hence , by corollary [ admissibled ] , the projective structure @xmath918 on @xmath55 is admissible .
since @xmath919 and @xmath910 is a convex subset of @xmath53 bounded by geodesics , by lemma [ c(i)/stc(l ) ] , @xmath920 embeds into @xmath921 . each boundary leaf of @xmath253
does not transversally intersect a leaf of @xmath47 with positive weight and @xmath253 is not a geodesic .
therefore , by corollary [ basiccoro ] , @xmath922 , where @xmath923 is the collapsing map for @xmath918 .
since @xmath918 is admissible , @xmath924 is also admissible . through its action
, @xmath925 is regarded as a schottky group in @xmath278 .
let @xmath926 be the bending map induced by @xmath927 .
then , by theorem [ ib ] , @xmath928 is an injective quasiisometric embedding , and it extends to an equivariant embedding of @xmath716 to @xmath8 . in particular , this extension takes the limit set of @xmath929 to the limit set of @xmath930 homeomorphically and @xmath931-equivariantly .
therefore , @xmath932 is an isomorphism onto a schottky group in @xmath9 .
hence , the restriction of @xmath1 to @xmath906 is admissible .
we have given a desired decomposition of @xmath2 , except that the flat cylinders @xmath933 are not integral .
in what follows , instead of cutting out the whole @xmath933 from @xmath0 , we cut out a maximal integral flat cylinder contained in @xmath933 . taking the union of the maximal integral flat cylinders and @xmath934
, we shall show that the complementary regions of the union are admissible , which completes the decomposition . for @xmath935 ,
let @xmath936 = \max \ { n \in \mathbb{z}_{\geq 0 } \>|\ > 2\pi n \leq x\}$ ] .
besides , let @xmath937 } { 2}\ ] ] for each @xmath876 . then @xmath938 . cut each @xmath939 along two admissible loops into three flat cylinders of heights @xmath940 , @xmath941 $ ] , @xmath940 in this order .
the middle cylinder @xmath942 is integral and the others are not .
( see @xmath486 in figure [ pandr ] . ) if @xmath943 , then @xmath942 degenerates to an admissible loop in the middle of @xmath939 ( see @xmath944 in figure [ pandr ] ) .
let @xmath945 and @xmath946 . each non - integral flat cylinder obtained above shares exactly one boundary component with a component of @xmath947 .
therefore , each component @xmath906 of @xmath907 is contained in a component @xmath858 of @xmath948 . if a boundary circle @xmath23 of @xmath906 maps to a periodic leaf @xmath949 via @xmath286 , then @xmath23 bounds a flat cylinder of height @xmath940 in @xmath858 . if @xmath23 maps to an approximating loop
@xmath794 by @xmath286 , then @xmath23 is a boundary component of @xmath858 .
thus , each component @xmath858 of @xmath950 is the union of a component @xmath906 of @xmath907 and the non - integral flat cylinders sharing a boundary component with @xmath906 . in particular
, @xmath906 is a deformation retract of @xmath858 .
letting @xmath951 be these non - integral flat cylinders , set @xmath952 ( the shaded region in figure [ pandr ] ) .
then @xmath953 , @xmath954 @xmath955 are periodic leaves of @xmath47 .
we have @xmath956 .
let @xmath957 be lifts of @xmath906 and @xmath858 to @xmath30 , respectively .
then @xmath958 , which is a lift of @xmath253 to @xmath53 .
let @xmath959 be the geodesic lamination on @xmath53 consisting of the boundary geodesics of @xmath910 that are lifts of periodic leaves of @xmath47 .
let @xmath960 ; then @xmath961 .
assign the weight @xmath99 to each leaf of @xmath962 , and obtain a measured lamination @xmath963 on @xmath53 .
recall that @xmath964 .
since there are no leaves of @xmath47 intersecting both @xmath910 and @xmath965 , and @xmath963 consists of isolated leaves of @xmath56 , therefore , @xmath966 and @xmath967 are disjoint
. then let @xmath968 .
each leaf @xmath23 of @xmath963 is a boundary geodesic of @xmath910 , and each leaf of @xmath927 intersects @xmath910 but does not intersect leaves of @xmath963 .
therefore , each @xmath23 is an outermost leaf of @xmath969 .
we now apply theorem [ ib2 ] with @xmath970 and @xmath971 .
we just have checked that @xmath972 ( hypothesis ( i ) of theorem [ ib2 ] ) .
every leaf of @xmath963 has the weight @xmath973 ( hypothesis ( iv ) ) .
since @xmath974 are disjoint simple loops on @xmath0 and @xmath963 consists of lifts of these loops , there exists @xmath509 such that every pair of distinct leaves of @xmath975 has a distance greater than @xmath640 ( hypothesis ( iii ) )
. apply theorem [ ib2 ] to this @xmath640 , and obtain @xmath976 .
apply proposition [ propid ] with @xmath977 , and obtain @xmath800 .
let @xmath179 be the component of @xmath869 containing @xmath253 .
then @xmath978 , where @xmath870 is a lift of @xmath179 to @xmath53 ( that contains @xmath910 ) . therefore , by the proposition [ propid ] , if @xmath801 , then @xmath979 ( hypothesis ( ii ) ) . by theorem [ ib2 ] ,
if @xmath37 is sufficiently large , @xmath980 is an injective @xmath658-quasiisometric embedding for every component @xmath201 of @xmath948 . by corollary
[ admissibled ] , the projective structure @xmath981 on @xmath55 is admissible .
let @xmath923 and @xmath982 be the collapsing maps associated with @xmath918 and @xmath981 , respectively . since @xmath983 , by corollary [ basiccoro ] , we have @xmath984 .
therefore , @xmath985 is admissible . for @xmath986 ,
let @xmath987 be a lift of @xmath988 to @xmath53 that bounds @xmath910 .
let @xmath405 be the component of @xmath989 bounded by @xmath987 . then , since @xmath910 is open , @xmath405 is a closed half plane .
note that @xmath405 is uniquely determined by the choice of @xmath285 and the choice of the lift of @xmath988 .
observe that @xmath990 .
then @xmath991 is a crescent of angle @xmath514 , and @xmath992 is a crescent of angle @xmath99 .
therefore , @xmath993 is a crescent of angle @xmath994 , and it is a component of @xmath995 .
each component of @xmath996 is a lift @xmath997 of some @xmath998 to @xmath30 .
there is a lift @xmath999 of @xmath751 bounding @xmath997 and @xmath1000 in @xmath330 ( figure [ randb ] ) .
the height of @xmath997 is @xmath1001 . by the argument above ,
when @xmath1002 is embedded in @xmath981 , @xmath1003 bounds a component of @xmath1004 , which is a crescent of angle @xmath994 .
therefore , this embedding extends to the embedding of @xmath1005 to @xmath981 ( see figure [ randb ] ) .
different components of @xmath1006 correspond to different @xmath997 .
therefore , the embedding disjointly extends to all components of @xmath1006 and we obtain an embedding of @xmath1007 into @xmath981 . by the construction ,
the embedding of @xmath1008 is @xmath1009-equivariant .
since @xmath1007 embeds to @xmath981 , and @xmath981 has an injective developing map , therefore @xmath1010 has an injective developing map .
since @xmath906 is a deformation retract of @xmath858 , @xmath1011 equals to @xmath1012 as subgroups of @xmath33 .
in particular , @xmath1013 . we have already seen that @xmath1009 is isomorphic to a purely loxodromic subgroup of @xmath9 via @xmath15 .
therefore , so is @xmath1014 .
hence , the restriction of @xmath1 to @xmath858 is admissible .
d. epstein and a. marden , _ convex hulls in hyperbolic space , a theorem of sullivan , and measured pleated surfaces _ , fundamentals of hyperbolic geometry , 117 - 266 , london math .
lecture note ser .
* 328 * , cambridge univ .
press , 2006 . | let @xmath0 be a closed orientable surface of genus at least two , and let @xmath1 be an arbitrary ( complex ) projective structure on @xmath0 . we show that there is a decomposition of @xmath0 into pairs of pants and cylinders such that the restriction of @xmath1 to each component has an injective developing map and a discrete and faithful holonomy representation .
this decomposition implies that every projective structure can be obtained by the construction of gallo , kapovich , and marden . along the way
, we show that there is an admissible loop on @xmath2 , along which a grafting can be done . |
bracket structures play an important role in classical differential geometry ( see , for example , refs .
@xcite ) , poisson geometry ( e.g. , @xcite ) , and the theory of integrable systems ( refs . @xcite ) .
being initially of a geometrical nature , these brackets found exact counterparts in abstract algebra , in the framework of vinogradov s theory of algebraic differential operators , @xcite ( see also book @xcite ) .
it became clear that many geometrical constructions ( such as the ones we meet in hamiltonian mechanics or in partial differential equations ; cf . with refs .
@xcite and @xcite , respectively ) may be more or less exactly expressed using the language of commutative algebra . in this paper
, i collected together my old results on the algebraic theory of the frlicher nijenhuis and schouten nijenhuis brackets and related homological and cohomological theories ( for shortness , i call these brackets nijenhuis and schouten ones ) .
these results were initially published in papers @xcite .
the results exposed below are easily generalized to the case of super - commutative algebras ( see @xcite ) and , being slightly modified , can be incorporated in lychagin s `` colored calculus '' ( see ref.@xcite ) . to simplify exposition
, i shall always assume that the algebra @xmath0 is such that the module @xmath1 ( see sec .
[ sec : algebr - calc - diff-1 ] ) of @xmath2-forms is projective and of finite type .
this scheme was first presented in ref .
@xcite .
let @xmath3 be a field , @xmath4 .
let also @xmath5 , @xmath6 be graded vector spaces and @xmath7 be endowed with a differential @xmath8 assume that there exists a graded monomorphism @xmath9 and define the `` _ _ lie derivative _ _ '' @xmath10,\qquad p\in\mathbb{p}.\ ] ] here and everywhere below @xmath11 $ ] denotes the graded commutator . if we are lucky then for two elements @xmath12 , @xmath13 we can define their _
@xmath14-bracket _ by @xmath15\!]}}^{\varphi}=[{\mathrm{l}}_p^{\varphi},{\mathrm{l}}_{p'}^{\varphi}],\qquad { [ \![}p , p'{]\!]}\in\mathbb{p}_{\alpha+\alpha'+\beta+1}.\ ] ] in some interesting cases we are lucky indeed .
let us introduce the basic notions of the calculus over commutative algebras that will be needed below ( see refs.@xcite for details ) .
consider a unitary commutative associative @xmath3-algebra @xmath0 and @xmath0-modules @xmath16 and @xmath17 .
[ sec : diff - oper-1 ] a @xmath3-linear map @xmath18 is a _ differential operator _ ( do ) of order @xmath19 if @xmath20\dots]]]=0\ ] ] for all @xmath21 .
the set of all dos @xmath22 forms an @xmath0-bimodule denoted by @xmath23 .
an operator @xmath24 is called a _ @xmath16-valued derivation _ if @xmath25 the module of these derivations is denoted by @xmath26 . define by induction the modules @xmath27 and set formally @xmath28 .
elements of @xmath29 are called _ multiderivations_. [ sec : diff - oper ] let @xmath30 be a smooth manifold and @xmath31 .
let also @xmath32 and @xmath33 be vector bundles over @xmath30 and @xmath34 , @xmath35 be the modules of their smooth sections .
then definition [ sec : diff - oper-1 ] gives the classical notion of a linear differential operator .
the correspondence @xmath36 is a representable functor from the category of @xmath0-modules to itself .
the corresponding representative object is denoted by @xmath37 and called the module of _ differential @xmath38-forms _ of the algebra @xmath0 .
in particular there exists a natural derivation @xmath39 such that any derivation @xmath40 uniquely decomposes as @xmath41 where @xmath42 .
the module @xmath43 is generated by the elements of the form @xmath44 @xmath45 while @xmath46 are exterior powers of @xmath43 .
this leads to the complex @xmath47&a\ar[r]^d&\lambda^1\ar[r]&\dots\ar[r ] & \lambda^i\ar[r]^d&\lambda^{i+1}\ar[r]&\dots } \ ] ] called the _ de rham _ complex of @xmath0 . due to the above formulated proposition , the module @xmath48 is a grassmannian algebra with the _ exterior _ , or _ wedge product _ @xmath49 a similar operation @xmath50 is introduced by induction in @xmath51 .
namely , for @xmath52 we set @xmath53 and @xmath54 @xmath55 , @xmath56 . in this way
, @xmath57 becomes a grassmannian algebra , @xmath58 being a module over @xmath57 . the _ inner product _
@xmath59 is defined by induction .
if @xmath60 we set @xmath61 and for @xmath62 we set @xmath63 if @xmath64 we set @xmath65 we shall use the notation @xmath66 when @xmath67 and @xmath68 , the inner product may be completed to the following operation @xmath69^{{\mathrm{i } } } & \lambda^k\otimes_a\lambda^{j - i}\ar[r]^{\wedge}&\lambda^{k+j - i } } \ ] ] which will be also called the inner product .
let @xmath70 and @xmath71 . then the maps @xmath72 and @xmath73 are super - differential operators of order @xmath38 and @xmath74 , respectfully .
we define here the schouten bracket , describe its properties and related ( co)homologies . some applications are also discussed .
let @xmath55 .
consider the lie derivative @xmath75\colon\lambda^j\to\lambda^{j - i}.\ ] ] for any two elements @xmath55 and @xmath76 there exists a uniquely defined element @xmath77\!]}^{\mathrm{s}}}\in d_{i+i'-1}(a)$ ] such that @xmath78={\mathrm{l}}_{{{[\![}{x},{x'}{]\!]}^{\mathrm{s}}}}.\ ] ] this element is called the _ schouten bracket _ of @xmath79 and @xmath80 .
we establish existence of @xmath77\!]}^{\mathrm{s}}}$ ] by induction .
for @xmath81 we set @xmath82\!]}^{\mathrm{s}}}=x(a),\qquad a\in a = d_0(a),\ ] ] and similarly for @xmath60 @xmath83\!]}^{\mathrm{s}}}=(-1)^{i'}x'(a).\ ] ] if @xmath84 we set @xmath85\!]}^{\mathrm{s}}}(a)={{[\![}{x},{x'(a)}{]\!]}^{\mathrm{s}}}+(-1)^{i'-1}{{[\![}{x(a)},{x'}{]\!]}^{\mathrm{s}}}.\ ] ] it is easily checked that thus defined bracket enjoys the needed property .
let @xmath86 be multiderivations of degree @xmath38 @xmath87 and @xmath88 respectively
. then 1 .
@xmath77\!]}^{\mathrm{s}}}+(-1)^{(i-1)(i'-1)}{{[\![}{x'},{x}{]\!]}^{\mathrm{s}}}=0 $ ] 2 .
@xmath89\!]}^{\mathrm{s}}}}{]\!]}^{\mathrm{s}}}={{[\![}{{{[\![}{x},{x'}{]\!]}^{\mathrm{s}}}},{x''}{]\!]}^{\mathrm{s } } } + ( -1)^{(i-1)(i'-1)}{{[\![}{x'},{{{[\![}{x},{x''}{]\!]}^{\mathrm{s}}}}{]\!]}^{\mathrm{s}}}$ ] 3 .
@xmath90\!]}^{\mathrm{s}}}= { { [ \![}{x},{x'}{]\!]}^{\mathrm{s}}}\wedge x''+(-1)^{(i-1)i'}x'\wedge{{[\![}{x},{x''}{]\!]}^{\mathrm{s}}}$ ] 4 .
@xmath77\!]}^{\mathrm{s}}}=[x , x']$ ] if @xmath91 5 .
@xmath92\!]}^{\mathrm{s}}}}=[{\mathrm{l}}_x,{\mathrm{i}}_{x'}]$ ] .
to any bivector @xmath93 , one can put into correspondence a skew - symmetric bracket @xmath94 , @xmath95 .
the following statements are equivalent 1 .
@xmath96 satisfies the jacobi identity 2 .
@xmath97\!]}^{\mathrm{s}}}=0 $ ] 3 .
@xmath98 where @xmath99\!]}^{\mathrm{s}}}$ ] .
if one of the previous conditions fulfills then : 1 .
@xmath100 is called a _
poisson structure _ and a pair @xmath101 is a _
poisson algebra_. 2 .
@xmath102 is the _ poisson bracket _ associated with @xmath100 .
3 . @xmath103 are _ hamiltonian derivations_. 4 .
derivations @xmath79 satisfying @xmath104 are _ canonical derivations_. let @xmath106 denote the algebra of all dos @xmath107 . for any @xmath108 the coset @xmath109_k=\delta\bmod\operatorname{diff}_{k-1}(a)$ ] is called its _ symbol_. if @xmath110_{k_1}$ ] , @xmath111_{k_2}$ ] are two symbols we define their product by @xmath112_{k_1+k_2}\ ] ] and their bracket by @xmath113_{k_1+k_2 - 1}.\ ] ] in such a way we obtain the _ algebra of symbols _ @xmath114 the above introduced algebra of symbols @xmath115 is a graded commutative algebra with a graded poisson bracket @xmath116 . in the case
@xmath31 it coincides with the algebra of smooth functions on @xmath117 polynomial along the fibers while the bracket is the one defined by the canonical symplectic form @xmath118 .
[ sec : example :- algebraic - t-1 ] the parallel between geometrical constructions and the corresponding algebraic modules is even deeper , though perhaps not so straightforward .
as an example , let us describe how the canonical form @xmath119 is defined within the model under consideration ( other illustrations can be found , e.g. , in refs .
@xcite ) .
note first that exactly in the same way as it was done above one can define symbols of arbitrary operators @xmath120 .
moreover , under the assumption of sec .
[ sec : introduction ] one has the isomorphism @xmath121 now , to define a @xmath2-form , one needs to evaluate all derivations on this form .
let @xmath122 be such a derivation , @xmath123 being an @xmath115-module . since @xmath124 , one can consider the restriction @xmath125 .
due to eq .
, one has @xmath126\in s_*(a , r)=s_*(a)\otimes_a r,\ ] ] and we set @xmath127 where @xmath128 is the multiplication .
consequently , we can define the form @xmath129 , but for general algebras it may be degenerate , contrary to the geometric case .
[ sec : example :- algebraic - t-2 ] it may be also appropriate to discuss another parallel here .
namely , in geometry , @xmath2-forms are identified with sections of the cotangent bundle . in algebra , the notion of section transforms to a homomorphism @xmath130 such that @xmath131 .
consider a @xmath2-form @xmath132 and let us define the corresponding `` section '' @xmath133 . to this end ,
recall that in the case when @xmath1 is a finite type projective module the algebra @xmath134 is generated by @xmath135 and @xmath136 .
therefore , any element @xmath137 is of the form @xmath138\cdot\ldots\cdot[x_{\alpha_k}]$ ] , where @xmath139 are derivations .
then we set @xmath140 conversely , let @xmath141 be a homomorphism . to define the corresponding @xmath2-form @xmath142
, we need to evaluate an arbitrary derivation @xmath24 at it , where @xmath16 is an @xmath0-module .
but @xmath143\in s_*(a , p)=s_*(a)\otimes_a p,\ ] ] and we set @xmath144)),\ ] ] where @xmath145 is the multiplication .
let @xmath101 be a poisson algebra .
the sequence @xmath146&a\ar[r]^{\partial_{\mathcal{p}}}&d_1(a)\ar[r]&\dots\ar[r ] & d_i(p)\ar[r]^{\partial_{\mathcal{p}}}&d_{i+1}(p)\ar[r]&\dots , } \ ] ] where @xmath99\!]}^{\mathrm{s}}}$ ] , is the _ poisson complex _ of @xmath0 and its cohomologies @xmath147 are the _ poisson cohomologies_. 1 .
@xmath148 consists of casimirs of @xmath100 and coincides with the poisson center of @xmath0 .
2 . @xmath149 . where @xmath150 is
the space of canonical derivations and @xmath151 consists of the hamiltonian ones .
@xmath152 coincides with the set of classes of nontrivial infinitesimal deformations of the poisson structure @xmath100 .
4 . @xmath153 contains obstructions to prolongation of infinitesimal deformations up to formal ones .
take a poisson algebra @xmath101 and consider the operator @xmath154\colon \lambda^j\to\lambda^{j-1}.\ ] ] by definition of the poisson structure , one has @xmath155= [ { \mathrm{l}}_{\mathcal{p}},{\mathrm{l}}_{\mathcal{p}}]= { \mathrm{l}}_{{{[\![}{\mathcal{p}},{\mathcal{p}}{]\!]}^{\mathrm{s}}}}=0\ ] ] and one gets the complex @xmath156&\lambda^j\ar[r]^{d_{\mathcal{p}}}&\lambda^{j-1}\ar[r]&\dots\ar[r ] & \lambda^1\ar[r]^{d_{\mathcal{p}}}&a\ar[r]&0 , } \ ] ] whose homologies @xmath157 are called the _ poisson homologies _ of @xmath101 .
the action of @xmath158 is fully defined by the following two properties : @xmath159 and @xmath160 let @xmath161 be the ideal generated by hamiltonian derivations .
let @xmath162 be its powers .
then @xmath163 is a filtration that generates a spectral sequence for poisson cohomologies . in a dual way , the filtration @xmath164 where @xmath165 gives rise to a spectral sequence for poisson homologies
the poisson bracket defined by a poissonian bi - vector @xmath100 can be extended to a super - bracket on the grassmannian algebra @xmath166 . consider the differential @xmath167 and a form @xmath71 .
then we have the map @xmath168 \colon d_i(a)\to d_{i - j+1}(a).\ ] ] for any two forms @xmath71 @xmath169 the equality @xmath170\ ] ] uniquely determines a form @xmath171 which is called their _ poisson bracket_. the poisson bracket of forms enjoys the following properties : 1 .
@xmath172 2 .
@xmath173 3 .
@xmath174 4 .
@xmath175 5 .
@xmath176 6 .
@xmath177 $ ] . note that the first three properties may be taken for the constructive definition of the extended bracket .
two poisson structures @xmath100 and @xmath178 _ commute _ , or are _ compatible _ if @xmath179\!]}^{\mathrm{s}}}=0 $ ] .
this is equivalent to @xmath180 or to the fact that @xmath181 ( the _ poisson pencil _ ) is a poisson structure for all @xmath182 , @xmath183 .
the magri scheme ( see ref @xcite ) that establishes existence of infinite series of commuting conservation laws for _ bi - hamiltonian systems _ has an exact algebraic counterpart : let @xmath0 be an algebra with two commuting poisson structures @xmath184 and @xmath185 and assume that @xmath186 .
assume also that two elements @xmath187 @xmath188 are given such that @xmath189
. then 1 .
there exist elements @xmath190 satisfying @xmath191 2 .
all elements @xmath192 are in involution with respect to both poisson structures i.e. @xmath193 for all @xmath194 @xmath195 .
consider a form - valued derivation @xmath196 and the lie derivative @xmath197\colon\lambda^j\to\lambda^{k+j},\ ] ] where @xmath198 is defined by the composition @xmath199^{{\mathrm{i } } } & \lambda^k\otimes_a\lambda^{j-1}\ar[r]^{\wedge}&\lambda^{k+j-1}. } \ ] ] the above lie derivative possesses the following properties 1 .
@xmath200 2 .
@xmath201=0 $ ] 3 .
@xmath202 . here
@xmath71 @xmath169 @xmath196 .
other basic properties of the nijenhuis bracket are presented in the following let @xmath203 @xmath204 @xmath205 and @xmath206
. then 1 .
@xmath207\!]}^{\mathrm{n}}}+(-1)^{jj'}{{[\![}{\omega'},{\omega}{]\!]}^{\mathrm{n}}}=0 $ ] + 2 .
@xmath208\!]}^{\mathrm{n}}}}{]\!]}^{\mathrm{n}}}= { { [ \![}{{{[\![}{\omega},{\omega'}{]\!]}^{\mathrm{n}}}},{\omega''}{]\!]}^{\mathrm{n}}}+ ( -1)^{jj'}{{[\![}{\omega'},{{{[\![}{\omega},{\omega''}{]\!]}^{\mathrm{n}}}}{]\!]}^{\mathrm{n}}}$ ] + 3 .
@xmath209\!]}^{\mathrm{n}}}={\mathrm{l}}_\omega(\omega)\wedge\omega'+ ( -1)^{ij}\omega\wedge{{[\![}{\omega},{\omega'}{]\!]}^{\mathrm{n}}}- ( -1)^{(j+1)(i+j')}d\omega\wedge{\mathrm{i}}_{\omega'}(\omega)$ ] + 4 . @xmath210=(-1)^j{\mathrm{l}}_{{\mathrm{i}}_{\omega'}}\omega+ { \mathrm{i}}_{{{[\![}{\omega},{\omega'}{]\!]}^{\mathrm{n}}}}$ ] + 5 .
@xmath211\!]}^{\mathrm{n}}}={{[\![}{{\mathrm{i}}_\omega\omega'},{\omega''}{]\!]}^{\mathrm{n}}}+ ( -1)^{(j+1)j'}{{[\![}{\omega'},{{\mathrm{i}}_\omega\omega''}{]\!]}^{\mathrm{n } } } + ( -1)^{j'}{\mathrm{i}}_{{{[\![}{\omega},{\omega'}{]\!]}^{\mathrm{n}}}}\omega''\\- ( -1)^{(j'+1)j''}{\mathrm{i}}_{{{[\![}{\omega},{\omega''}{]\!]}^{\mathrm{n}}}}\omega'$ ] . on decomposable elements the nijenhuis bracket acts as follows .
let @xmath71 , @xmath169 , @xmath79 , @xmath212
. then @xmath213\!]}^{\mathrm{n } } } & = \omega\wedge\omega'\wedge[x , x']+{\mathrm{l}}_{\omega\wedge x}(\omega')\wedge x'- ( -1)^{jj'}{\mathrm{l}}_{\omega'\wedge x'}(\omega)\wedge x\\ & = \omega\wedge\omega'\wedge[x , x']+\omega\wedge{\mathrm{l}}_x(\omega')\wedge x'- { \mathrm{l}}_{x'}(\omega)\wedge\omega'\wedge x\\ & + ( -1)^jd\omega\wedge{\mathrm{i}}_x(\omega')\wedge x ' + ( -1)^j{\mathrm{i}}_{x'}(\omega)\wedge d\omega'\wedge x.\end{aligned}\ ] ] a derivation @xmath214 is called _ integrable _ if @xmath215\!]}^{\mathrm{n}}}=0.\ ] ] with any integrable derivation one can associate a complex @xmath146&d_1(b)\ar[r]^{\partial_{\mathcal{n}}}&\dots\ar[r ] & d_1(\lambda^j(b))\ar[r]^{\partial_{\mathcal{n}}}&d_1(\lambda^{j+1}(b))\ar[r]&\dots } \ ] ] where @xmath216\!]}^{\mathrm{n}}}$ ] . such structures , in particular ,
arise in an algebraic model of flat connections .
let @xmath0 and @xmath217 be @xmath3-algebras and @xmath218 be a homomorphism .
then @xmath217 is an @xmath0-algebra and one can consider the module @xmath219 of @xmath217-valued derivations @xmath220 . for any @xmath221 denote by @xmath222 its restriction to @xmath0 .
a _ connection _ is a @xmath217-homomorphism @xmath223 such that @xmath224 . a vector - valued form @xmath225 defined by @xmath226
is called the _ connection form_. for any two derivations @xmath79 , @xmath227 we set @xmath228-\nabla(\nabla(x)\circ x ' -\nabla(x')\circ x);\ ] ] @xmath229 is the _ curvature _ of @xmath230 .
a connection is _ flat _ if @xmath231 .
let @xmath230 be a connection
. then @xmath232\!]}^{\mathrm{n}}}))=2r_\nabla({\left.x\right|}_a,{\left.x'\right|}_a)\ ] ] for any @xmath79 @xmath233 . hence , to any flat connection
, i.e. , to a connection whose curvature vanishes , we associate a nijenhuis complex with @xmath234 . in applications ,
its _ vertical _ subcomplex is useful : @xmath146&d_1^v(b)\ar[r]^{\partial_{\mathcal{n}}}&\dots\ar[r ] & d_1^v(\lambda^j(b))\ar[r]^{\partial_{\mathcal{n } } } & d_1^v(\lambda^{j+1}(b))\ar[r]&\dots , } \ ] ] where @xmath235 .
denote its cohomologies by @xmath236 .
let @xmath230 be a flat connection
. then 1 .
the cohomology groups @xmath236 inherit the inner product operation @xmath240 in particular the group @xmath241 is an associative algebra represented in endomorphisms of @xmath242 @xmath243 2 .
the cohomology groups @xmath236 inherit the nijenhuis bracket @xmath244\!]}^{\mathrm{n}}}\colon h^j(b,\nabla)\times h^{j'}(b,\nabla)\to h^{j+j'}(b,\nabla).\ ] ] in particular @xmath242 is a lie algebra @xmath244\!]}^{\mathrm{n}}}\colon h^0(b,\nabla)\times h^0(b,\nabla)\to h^0(b,\nabla).\ ] ] let @xmath245 be an infinitely prolonged differential equation in the jet bundle of a bundle @xmath246 .
the bundle @xmath247 is always endowed with a natural flat connection @xmath248 ( the _ cartan connection _ , see refs @xcite ) and taking @xmath249 we obtain the picture considered above .
let us use the notation @xmath250 for the cohomology groups arising in this case . for any formally integrable equation @xmath251 that surjectively projects to @xmath252 one has 1 .
@xmath253 coincides with the lie algebra @xmath254 of higher symmetries of @xmath251 .
elements of @xmath255 act on @xmath254 and thus are identified with recursion operators for symmetries .
3 . on the other hand elements of @xmath255 can be understood as classes of nontrivial infinitesimal deformations of the equation structure .
4 . @xmath256 contains obstructions to prolongation of infinitesimal deformations up to formal ones .
let @xmath257 be an algebra with flat connection .
for @xmath258 and @xmath259 , use the notation @xmath260 , @xmath261 , @xmath262 assume that @xmath263 .
then for any @xmath79 @xmath264 and @xmath259 for all @xmath265 @xmath266 one has @xmath267=[x , y]_{m+n}+\sum_{i=0}^{n-1}({{[\![}{x},{r}{]\!]}^{\mathrm{n}}}(y_i))_{m+n - i-1}- \sum_{j=0}^{m-1}({{[\![}{y},{r}{]\!]}^{\mathrm{n}}}(x_j))_{m+n - j-1}.\ ] ] if @xmath268\!]}^{\mathrm{n}}}={{[\![}{y},{r}{]\!]}^{\mathrm{n}}}=0 $ ] and @xmath269=0 $ ] then @xmath270=0 $ ] for all @xmath265 @xmath266 .
if @xmath251 is a scalar evolutionary equation of order @xmath271 then @xmath272 .
let @xmath214 be an integrable element , i.e. , @xmath273\!]}^{\mathrm{n}}}=0 $ ] .
then the operator @xmath274 is a differential : @xmath275 .
moreover , one has @xmath276=0\ ] ] and consequently the pair @xmath277 , where @xmath278 , constitutes a bi - complex that converges to the de rham cohomologies of @xmath217 . in the case of differential equations
( @xmath31 , @xmath279 , and @xmath280 is the connection form of the cartan connection in the bundle @xmath281 ) , this bi - complex coincides with the _ variational bi - complex _ , or _
vinogradov s @xmath248-spectral sequence _ , see ref .
to conclude , note that several more brackets can be constructed in a similar way . 1 .
first , mention the _ nijenhuis
richardson bracket _
@xmath244\!]}^{\mathrm{r}}}\colon d_1(\lambda^j)\otimes_a d_1(\lambda^{j'})\to d_1(\lambda^{j+j'-1})\ ] ] that can be defined by @xmath282={\mathrm{i}}_{{{[\![}{\omega},{\omega'}{]\!]}^{\mathrm{r}}}}\ ] ] and is of the form @xmath283\!]}^{\mathrm{r}}}={\mathrm{i}}_\omega(\omega')- ( -1)^{(i-1)(j-1)}{\mathrm{i}}_{\omega'}(\omega)\ ] ] and is one of the classical and well known brackets .
two more brackets arise also if we fix a poisson structure @xmath284 or a nijenhuis structure @xmath285 : 1 .
consider the inner product @xmath286 then for @xmath287 the following `` lie derivative '' arises : @xmath288\colon d_k(a)\to d_{k+i}(a)\ ] ] and one can introduce a bracket @xmath289\!]}^{\mathcal{p}}\colon d_i(\lambda^1)\times d_{i'}(\lambda^1)\to d_{i+i'}(\lambda^1)\ ] ] by @xmath290\!]}^{\mathcal{p}}}^{\mathcal{p}}= [ { \mathrm{l}}_\omega^{\mathcal{p}},{\mathrm{l}}_{\omega'}^{\mathcal{p}}].\ ] ] 1 . in a similar way , one can consider the inner product @xmath291 and the `` lie derivative '' @xmath292\colon d_1(\lambda^k)\to d_1(\lambda^{k+i}),\ ] ] @xmath293 . then
a new bracket @xmath289\!]}^{\mathcal{n}}\colon d_1(\lambda^i)\times d_1(\lambda^{i'})\to d_1(\lambda^{i+i'})\ ] ] is defined by @xmath290\!]}^{\mathcal{n}}}^{\mathcal{n}}= [ { \mathrm{l}}_\omega^{\mathcal{n}},{\mathrm{l}}_{\omega}^{\mathcal{n}}].\ ] ]
one can also define the inner products @xmath294 and @xmath295 together with the corresponding lie actions @xmath197\colon\lambda^k\to\lambda^{k+j - i+1}\ ] ] and @xmath288\colon d_k(a)\to d_{k - j+i+1}(a),\ ] ] where @xmath296 .
of course , it is tempting to find the elements @xmath297\!]}$ ] and @xmath297\!]}^{\mathcal{p}}$ ] such that @xmath290\!]}}=[{\mathrm{l}}_\omega,{\mathrm{l}}_{\omega'}],\qquad { \mathrm{l}}_{{[\![}\omega,\omega'{]\!]}^{\mathcal{p}}}^{\mathcal{p}}= [ { \mathrm{l}}_\omega^{\mathcal{p}},{\mathrm{l}}_{\omega'}^{\mathcal{p}}],\ ] ] but in general such elements do not exist ( see discussion of these matters in ref .
@xcite ) .
bocharov , a.v .
, chetverikov , v.n . , duzhin , s.v . , et al . , symmetries and conservation laws for differential equations of mathematical physics , xiv+333 pp . amer .
soc . , providence , ri ( 1999 ) . edited and with a preface by i. krasilshchik and a. vinogradov .
i. krasilshchik , algebras with flat connections and symmetries of differential equations , in : _ lie groups and lie algebras : their representations , generalizations and applications _ , kluwer acad .
publ . , dordrecht / boston / london , 1998 , 407424 .
i. krasilshchik , schouten brackets and canonical algebras , _ global anal . and math . physics
, voronezh state university , 1987 .
english translation in springer _ lecture notes in math .
_ , * 1334 * , springer - verlag , 1988 .
i. krasilshchik , v. lychagin , a. vinogradov , geometry of jet spaces and nonlinear differential equations , _ advanced studies in contemporary mathematics _ ,
* 1 * ( 1986 ) , gordon and breach , new york , london . xx+441 | a general theory of the frlicher nijenhuis and schouten nijenhuis brackets in the category of modules over a commutative algebra is described . some related structures and (
co)homology invariants are discussed , as well as applications to geometry . |
the discovery of a kink " in the nodal ( ( 0,0 ) - ( @xmath1,@xmath1 ) ) dispersion of the high - t@xmath2 cuprates , and band renormalizations , in the form of a peak - dip - hump structure in the anti - nodal ( 0,@xmath1 ) - ( @xmath1,@xmath1 ) dispersion @xcite , have attracted considerable attention in recent years . these band renormalizations are interpreted as due to electron - boson coupling , and it is believed that understanding their origin will provide information about the underlying pairing mechanism in these materials . there is still considerable debate as to the identity of the responsible bosonic mode@xcite .
one proposal is coupling to an electronic mode associated with the collective mode found in neutron scattering near momentum transfers ( @xmath1,@xmath1 ) , the so called magnetic resonance mode @xcite .
additionally , renormalizations could also be due to coupling of electrons to damped magnons , which have less well - defined momentum structure .
however , the strength of the coupling of magnons to electrons is under debate@xcite .
for example , quantitative comparisons of arpes and neutron measurements on yba@xmath0cu@xmath3o@xmath4 ( ybco ) have been made , and the overall strength of the coupling inferred from the data was indicated to be of sufficient strength to give rise to superconductivity@xcite .
we also remark that a quantitative comparison between the neutron scattering and arpes measurements reported in ref . can be complicated by the polar surface of cleaved ybco resulting in surface reconstruction .
this can produce significant differences between the bulk and surface layers of this material .
moreover , a systematic examination of the effects of doping , material class , and temperature have not been thoroughly explored .
in fact , given that the spin continuum arises from the cuo@xmath0 plane , one might expect the coupling to electrons to be relatively material - class independent .
on the other hand , it is well - known that a neutron resonance displays a material dependence , appearing at larger energies for larger t@xmath2 materials including both the single and multi - layer cuprates .
this opens the possibility of linking the neutron resonance with arpes renormalizations via a material - dependent study .
an alternative proposal is coupling to a spectrum of oxygen vibrational phonon modes @xcite , specifically , the @xmath5-axis out - of - phase bond - buckling oxygen vibration or b@xmath6 mode ( @xmath7 mev ) and the in - plane bond - stretching oxygen mode ( @xmath8 mev ) .
this proposal has been able to account for many experimental observations including the anisotropy of the observed renormalizations@xcite , fine structure in the form of subkinks observed in the temperature dependence of the self - energy in bi@xmath0sr@xmath0cacu@xmath0o@xmath9@xcite ( bi-2212 ) , and doping dependent changes in self - energy @xcite .
this interpretation is further supported by recent arpes experiments that have measured an @xmath10 @xmath11 @xmath12 isotope shift in the nodal kink position@xcite .
finally , analogous features have been observed in tunneling spectra which have also been interpreted as coupling to a bosonic mode @xcite , and possibly the same mode responsible for the arpes observed kink .
ref . reported an @xmath13o isotope shift of the feature in a bi-2212 sample indicating a phonon origin . in this work
, we present a review of how these two scenarios can be differentiated by studying the material , doping and temperature dependence of the band renormalizations .
our focus is on how @xmath5-axis phonons provide a material dependence to the arpes kinks due to the phonon s sensitivity to local symmetry and the environment surrounding the cuo@xmath0 plane .
we consider doping dependent changes to the renormalization in bi-2212 as well as the dependence of the renormalization within the bi and tl families as the number of cuo@xmath0 layers is varied .
we will also discuss some recent arpes results on the @xmath14 layer system ba@xmath0ca@xmath3cu@xmath15o@xmath16f@xmath0 ( f0234 ) @xcite . in this system ,
the inner and outer layers occupy different crystal environments resulting in differing madelung energies associated with each plane in the undoped compound .
this difference drives inequivalent dopings between the two sets of layers , with one set @xmath17-type and the other @xmath18-type .
the inequivalent doping in each plane generates further symmetry breaking in the layers and the el - ph coupling is expected to differ .
indeed , ref . observes stronger kink features in the plane associated with the outer ( @xmath17-type ) layer of the material . here
, we will discuss these observations in the context of the el - ph coupling scenario .
the organization of this paper is follows . in section
ii we discuss the role of symmetry breaking in producing el - ph coupling to @xmath5-axis phonons and how such coupling is expected to vary with the crystal structure of the high - t@xmath2 cuprates . in section
iii we present arpes data for various multi - layer cuprates in order to examine how the now the band renormalizations vary with the number of cuo@xmath0 layers , @xmath17 .
energy distribution curves in the anti - nodal region of the bi and tl - families for @xmath19 and mdc dispersions for the @xmath18-bands of the f - family with @xmath20 are presented .
of particular interest are the results for the single layer tl cuprate tl-2201 . here
, we show that tl-2201 does not resolve the typical peak - dip - hump structures in the antinodal region despite the fact that the spin resonance mode exists in this system@xcite . in section iv the doping dependence of the nodal and antinodal dispersions for bi-2212 are presented .
the renormalizations in each region behave differently as the samples are overdoped , pointing to presence of multiple bosonic modes . in section
v we present a theoretical basis for understanding the self - doping phenomena in the parent compound f0234 and discuss what implications this process has on the coupling to @xmath5-axis phonons .
finally , in section vi , we conclude with a brief summary and some additional remarks .
in this section we discuss how @xmath5-axis phonons can be sensitive to the material environment off the cuo@xmath0 planes .
electron - phonon ( el - ph ) coupling to the in - plane bond - stretching modes are of a deformation type and the strength of the coupling depends on the cu - o bond distance .
since this distance is relatively constant in all cuprates , coupling to these modes is relatively materials independent .
this is not true for coupling to the @xmath5-axis bond buckling modes .
since our goal is to explore the materials dependence of the band renormalizations , we will focus our attention to the @xmath5-axis modes .
el - ph coupling to @xmath5-axis phonons can arise to first order from modulations in the on - site energy due to atomic site oscillation through local crystal fields .
for the cuo@xmath0 planes , such crystal fields arise due to asymmetries in the environment surrounding the plane @xcite and the degree of symmetry breaking varies from material to material with the chemical environment ( number of layers and doping from the ideal stoichiometric compound ) . in an ideally undoped single layer cuprate
the structure above and below the planes is identical and little symmetry breaking occurs across the plane .
the coupling to the planar @xmath5-axis modes is therefore expected to be weak in the parent compound . in the undoped multi - layer systems symmetry breaking across the outer planes can be large leading to a sizeable coupling in these layers .
a second pathway for symmetry breaking occurs as the parent compounds are doped by introducing substitutional or interstitial dopant atoms in the charge reservoir area off the cuo@xmath0 planes .
these dopants donate charge to the cuo@xmath0 plane , and cast off - plane electric fields @xmath21 that can not be effectively screened by the in - plane carriers@xcite . as a result they form charge impurities that further break symmetry across the cuo@xmath0 planes . through this mechanism , coupling to @xmath5-axis phonons can occur in single layer systems where they are normally forbidden by symmetry . the local crystal field strength at the planar oxygen site of the outermost cuo@xmath0 plane of the hg - family of cuprates .
all results have been normalized by the maximum field which occurs for the @xmath22 layer system . ] to quantify the effect of the crystal environment in the parent systems , we calculate the local crystal field at the outermost cuo@xmath0 layer for the hgba@xmath0ca@xmath23cu@xmath23o@xmath24 ( n = 1 - 6 ) family of cuprates . to do this ,
we use experimental structural data @xcite and assume an ionic point charge model with formal valences assigned to each atom . the ewald summation method @xcite is then used to perform the electrostatic sums for the electric field strength .
the results are shown in fig .
[ fig : efields ] . in the single layer compound the local field is zero since the cuo@xmath0 layer coincides with the plane of mirror symmetry and we have not considered dopings away from pure stoichiometry . in the double- and triple - layer compounds the value of the field rises taking a maximum for @xmath25 .
after this , as the number of layers further increases , there is a general trend of decreasing field values .
( we note here that the structural data for the @xmath26 and @xmath27 compounds had a large degree of error , presumably from the difficulty in sample growth . )
this decrease is due to a reduction in the degree of symmetry breaking across the outer plane once the number of layers is increased beyond @xmath25@xcite . in the limit of the infinite layer system
cacuo@xmath0 the environment is symmetric around every plane . in this case
the electrostatic contribution to the coupling is identically zero however steric forces introduce buckling to the plane which creates a new pathway for el - ph coupling of a deformation type@xcite .
however , in terms of material dependence , such steric forces are present in all cuo@xmath0 systems and therefore do not contribute to differences between materials .
the strength of the coupling to @xmath5-axis modes , such as the b@xmath6 mode , scales as @xmath28 .
furthermore , due to the oxygen charge transfer form factos , the b@xmath6 mode couples most strongly to antinodal electrons@xcite .
therefore the behavior of the antinodal renormalization is expected to have a dramatic materials dependence if it is due to the b@xmath6 phonon .
this layer dependence is expected to be different for coupling to the spin - resonance mode .
therefore , by examining the change in coupling strength as one moves up in the number of layers provides a pathway for distinguishing between these two scenarios .
in the previous section we discussed how the material dependence of the renormalizations is expected to arise in the various families of cuprates and how this coupling is expected to differ in the phonon and spin resonance proposals .
we now wish to review the available arpes data in light of the theoretical considerations of the previous section . here
, our focus is on the observed changes in the renormalizations as the number of layers within the bi- and tl - families .
single crystals of nearly optimally doped tl@xmath0ba@xmath0cacu@xmath0o@xmath16 ( tl-2212 ) , tlba@xmath0ca@xmath0cu@xmath3o@xmath29 ( tl-1223 ) and slightly overdoped tl@xmath0ba@xmath0cuo@xmath30 ( tl-2201 ) were grown using the flux method .
as - grown tl2212 ( t@xmath2 = 107 k ) and tl-1223 ( t@xmath2 123 k ) crystals were chosen for the arpes measurements .
tl-2201 crystals used in our measurement were prepared by annealing the as - grown crystal ( t@xmath2 @xmath31 30 k ) under a nitrogen flow at a temperature of 500@xmath32c , yielding a t@xmath2 of 80 k. the data were collected using a scienta r4000 photoelectron spectrometer .
measurements were performed at the ssrl beam line 5 - 4 using 28 ev photons and at the advanced light source beam line 10.0.1 using 50 ev photons .
the energy resolution was set at 15 - 20 mev for the tl data presented in this work .
samples were cleaved and measured in ultrahigh vacuum ( @xmath33 torr . ) to maintain a clean surface .
detailed arpes results on these compounds have been reported in ref . .
although the dispersion kink along the nodal direction has been found universally in high - t@xmath2 cuprates @xcite , the momentum dependence of this renormalization feature , when moving away from the nodal direction , exhibits a material dependence@xcite .
it has been confirmed recently that there is dependence on the number of cuo@xmath0 planes in the unit cell in the bi - family and , most recently , in the tl - family of cuprates@xcite . in the multi - layer compounds ,
the kink " becomes more dramatic and eventually breaks the band dispersion into two branches : one branch with a sharp peak and another branch with a broader hump structure @xcite .
the two branches asymptotically approach one another at a characteristic energy scale of 70 mev and coincide with the dominant energy scale of the kink along the nodal dispersion for nearly optimally - doped cuprates .
this separation of the band dispersion becomes most prominent near the antinodal region and results in the famous peak - dip - hump structure@xcite in the energy distribution curves ( edcs ) , as shown in fig .
[ fig : layers ] . 0.5 cm representative edcs near the antinodal region of the bi- and tl - families of cuprates , including single layer ( bi-2201 and tl-2201 ) , bi - layer ( bi-2212 and tl-2212 ) , and tri - layer ( bi-2223 and tl-1223 ) compounds .
the high background in the data of tl-1223 is probably due to the absence of a natural cleaving plane in the crystal structure .
nevertheless , a peak - dip - hump structure in the spectrum can still be discerned .
the red dashed line is a guide - to - the - eye to make the hump " more discernible .
, title="fig : " ] the momentum dependence of the kink is quite different in the single layer compound , where the dispersion kink becomes less prominent moving away from the node .
in addition , the band dispersion retains a single branch with no separation observed , unlike the case of the multi - layer compounds @xcite . as a result , no apparent peak - dip - hump structure can be seen in the edcs near the antinodal region for the single layer compounds fig .
[ fig : layers]@xcite . in summary ,
the layer dependent renormalization near the antinodal region is due most likely to electrons coupled to a sharp bosonic mode , whose origin is strictly constrained by the number of layers in the material . this mode is either absent , or has a negligible coupling to the electrons , in single layer compounds , but exhibits prominent coupling in the multi - layer compounds .
the spin resonance mode does exist in some single layer cuprates@xcite ( notably tl-2201 by not la@xmath0cuo@xmath34 ) .
therefore , one can conclude that the spin resonance mode is an unlikely candidate for the mode responsible for the renormalizations in the antinodal region . on the other hand , coupling to @xmath5-axis phonons can exhibit a very different coupling in single- and multi - layer compounds . as we have discussed , the b@xmath6 phonon couples strongly to the electrons in multi - layer compounds and weakly to electrons in single layer compounds .
this mode can also reproduce the observed anisotropic momentum dependence of the renormalization in bi - layer bi-2212@xcite .
we also note that the form for the b@xmath6 coupling is attractive in the @xmath35-wave pairing channel@xcite , which could be one factor enhancing t@xmath2 in the multi - layer systems .
we now consider another case , that of the f - family of cuprates with @xmath17 cuo@xmath0 layers , @xmath36 .
the single crystalline samples were grown by the flux method under high pressure@xcite .
arpes measurements on the f - family were performed at beamline 10.0.1 of the advance light source ( als ) at lawrence berkeley national laboratory .
the measurement pressure was kept @xmath37 torr at all time and data were recorded by scienta r4000 analyzers at 15k sample temperature .
the total convolved energy and angle resolution were 16 mev and 0.2@xmath32 respectively for photoelectrons generated by 55 ev photons . in fig .
[ fig : fl_mdc ] we present mdc derived dispersions for @xmath18-type bands of the three ( f0223 ) , four ( f0234 ) and five ( f0245 ) layer f - based cuprates . in all three cases
the dispersions show clear kinks , but at a larger energy scale in the four- and five - layer materials . in a @xmath35-wave superconductor coupled to an einstein mode , the energy scale of the kink occurs at @xmath38 where @xmath39 is the energy of the mode and @xmath40 is the maximum value of the superconducting gap .
this shift in energy scale is a reflection of the changes in the superconducting gap size as @xmath17 is varied from 3 to 5 . in order to quantify the strength of the kink
, slopes are extracted from the dispersion above and below the kink position , @xmath41 and @xmath42 .
an estimate for the relative coupling stengths @xmath43 is then given by : mdc derived dispersions along the nodal direction ( 0,0 ) - ( @xmath1,@xmath1 ) of the @xmath18-type band in the 3-layer ( f0223 ) , 4-layer ( f0234 ) and 5-layer ( f0245 ) f - family of cuprates . ]
@xmath44 this procedure produces @xmath45 , @xmath46 and @xmath47 for @xmath25 , @xmath48 and @xmath49 respectively .
this trend is easily understood in the phonon mode scenario where the dominant mode in the superconducting state is the b@xmath6 mode for which the coupling strength is proportional to the local crystal field .
the observed decrease in coupling strength can be easily understood from the expected local field strengths in the outer most layers ( recall fig .
[ fig : efields ] ) , which decreases for @xmath50 .
in the el - ph coupling picture the carriers couple to a spectrum of bosonic modes and we have already seen how the @xmath5-axis modes can produce a materials dependence of the renormalizations .
it is important to note that the coupling to each of these modes is highly anisotropic with the in - plane bond - stretching ( breathing ) mode dominating in the nodal region and the out - of - plane bond - buckling ( b@xmath6 ) mode dominating in the anti - nodal region . since these modes have different frequencies one would naturally expect the multiple energy scales to manifest in the experimental data .
indeed , evidence for multiple energy scales has been found both in the temperature dependence @xcite as well as the doping dependence of bi-2212 @xcite . in this section
we revisit the doping dependence of the nodal and antinodal renormalizations , highlighting the different behavior in each region of the brillouin zone and discuss how this dichotomy further supports the el - ph scenario .
high quality single crystals of optimally doped bi@xmath0sr@xmath0ca@xmath51y@xmath52cu@xmath0o@xmath9
( bi2212 op , t@xmath2 = 96 k ) were grown by the floating zone method .
the overdoped crystals with t@xmath2 = 88 k were prepared by post annealing the optimally doped bi2212 crystal under oxygen flow at a temperature of 400@xmath32c .
the overdoped sample with t@xmath2 = 65k is a derivative of the bi2212 family with lead doped into the crystal to achieve such an overdoped configuration .
the data were collected by using he i light ( 21.2 ev ) from a monochromated and modified gammadata he lamp with a scienta-2002 analyzer and in ssrl beamline 5 - 4 using 19 ev photons with a scienta-200 analyzer .
the energy resolution is @xmath3110 mev and angular resolution @xmath310.35@xmath32 .
the samples were cleaved and measured in ultra high vacuum ( @xmath53 torr . ) to maintain a clean surface . in fig .
[ fig : scaleantinode ] arpes data taken along a cut in the antinodal region illustrate this effect .
the upper panels show the measured spectral function along the cut while the lower panels show @xmath54 at a fixed k - point as indicated by the dashed lines .
the doping dependence of the antinodal spectrum of bi2212 taken in the superconducting state ( 10k ) . shown in the upper row are the false color plots of the spectra taken along the indicated cut direction ( inset ) .
the black dots are the peak and hump positions of the bonding band seen in the edcs . shown in the lower row
are the edcs along the dashed line indicated in the false color plots .
the symbols ab " and bb " represent the antibonding and binding bands while the numbers are the energy position of the dip of the edc . ] near ( 0,@xmath1 ) , the energy of the dip feature is the best measure of the energy scale of the mode responsible for the renormalization @xcite .
for the optimally doped sample ( op96k ) , shown in the first column of fig .
[ fig : scaleantinode ] , the dip position is clearly located at @xmath55 mev . this can be seen in both the false color plot and the edc cut . for moderate overdoping ( od88k )
, the energy of the dip is lowered to @xmath56 mev while for heavily overdoped ( od65k ) the energy is lowered further to @xmath57 mev . in both overdoped cases , contributions from the bonding- ( bb ) and anti - bonding ( ab ) bands contribute to the quasiparticle peak at the fermi level . in the od65k case , the contribution from the ab makes an exact determination of the dip position difficult and the estimate of @xmath57 mev should be considered a lower bound .
the doping dependence of the nodal spectrum at a temperature well below t@xmath2 ( 10 k ) . shown in the upper row are the false color plots of the spectra taken along the cut as indicated in the inset .
the black curves are band dispersion obtained by fitting mdcs to lorentizan functions .
the apparent kink position in the dispersion are marked by the yellow shaded area , which appears to be approximately the same for all three dopings .
the dashed lines serve as a guide - to - the - eye for visualizing the apparent kink in the dispersion . shown in the lower row
are the real part of the self - energy extracted from subtracting the band dispersion from a linear bare band .
the arrows indicate the positions of fine structure in the extracted re@xmath58 . ]
turning now to the nodal region , we find qualitatively different behavior . fig .
[ fig : scalenode ] presents @xmath54 along the nodal cut ( ( 0,0 ) - ( @xmath1,@xmath1 ) ) for the same three samples .
the highlighted region indicates the approximate position of the kink . in the nodal region
the overall bandwidth is much larger than the energy of the bosonic modes so the dramatic band breakup does not occur @xcite and the renormalization manifests as a kink in the dispersion . in this case , the energy scale of the kink is most easily determined from the structure of the real part of the self - energy re@xmath58 . the mdc - derived estimate for re@xmath58 , obtained from subtracting the mdc - derived dispersion from an assumed linear band , is also in the lower panels of fig .
[ fig : scalenode ] .
the nominal doping dependence @xcite of the energy scales in the nodal and antinodal region of bi2212 are summarized in fig .
[ fig : summary]a . for reference ,
the superconducting gap @xmath40 is shown also , which is determined from the peak positions of the fermi function divided spectrum at the fermi level .
while the characteristic energy in the antinodal region ( dip energy ) follows the decrease in the superconducting gap , the characteristic energy in the nodal region remains more or less constant ( @xmath59 mev ) .
the difference in the doping dependence of the two energy scales lends further support to the existence of coupling to multiple modes .
if a single mode were responsible for the renormalization throughout the zone one would expect the doping dependence to follow the same trend in the nodal and antinodal regions .
the energy of the dominant mode @xmath39 can be obtained by subtracting the magnitude of the superconducting gap from the observed energy scale , expected to be @xmath38@xcite .
the results of this procedure are shown in fig .
[ fig : summary]b .
the energy of the dominant mode in the antinodal region , within the error bars of the data , is independent of doping .
the behavior in the nodal region is different ; the energy of the dominant mode changes with doping . at optimal doping
the energy of the dominant mode is @xmath60 mev but in the overdoped samples the energy is larger @xmath61 mev .
this result is consistent with the picture of coupling to multiple modes outlined in ref . .
we also note that in od88k a secondary feature can be observed in re@xmath58 at precisely the same energy as the dip energy of the antinodal region .
similar fine structure was reported earlier in ref . .
the presence of this sub - feature in the ud88k data as the sample is progressively overdoped , along with the 35 mev scale in the nodal data at optimal doping , is evidence of a trade off between a coupling dominated by the b@xmath6 mode and one dominated by the bond stretching mode .
we further note that ref .
reached similar conclusions but assigned the anti - nodal renormalizations to the spin resonance mode .
we believe that the multi - layer data of the previous section , especially the single - layer tl data which shows no renormalization in the anti - nodal region , directly refutes this conclusion and favors the el - ph scenario . 0.5 cm a summary of some energy scales relevant to the renormalized band dispersions .
( a ) the apparent kink position in the nodal dispersion , dip energy at the antinodal region and the superconducting gap are summarized for the three doping levels shown in fig [ fig : scaleantinode ] .
( b ) the mode energy obtained by subtracting the superconducting gap from the characteristic energies of the renormalization effect .
the red arrow is to remind the reader that the shown quantity at the antinodal region of the od65k sample is a lower bound for the actual value.,title="fig : " ]
we now turn our attention to the identity of the carrier types in the inner and outer planes in f0234 .
experimentally , the parent compound of f0234 is known to self - dope with the inner and outer planes having fermi surfaces of different carrier type .
a recent arpes study@xcite found that the @xmath18-type bands are bilayer split along the nodal direction .
since the inner planes are expected to have a stronger inter - planar coupling than the outer planes , the observation of bilayer splitting provides strong evidence that the inner layers are @xmath18-type while the outer layers are @xmath17-type .
furthermore , kinks in the nodal dispersion of both sets of planes but the strength of the coupling in the @xmath17-type layer by a factor of two . a summary of the electrostatic calculations for f0234 .
( a ) the madelung potential difference between the cu ( red ) and o ( black ) sites of the inner and outer planes as a function of the oxygen buckling of the outer layer .
the shaded region indicates the region where the inner and outer planes are identified as @xmath18- and @xmath17-type respectively .
buckling of the planar o away from the mirror plane of the crystal is defined to be positive .
( b , inset ) the filling of the inner ( red ) and outer ( black ) planes as a function of the site energy difference ( buckling ) between inner and outer layers .
fillings are obtained with a simple tight - binding model ( see text ) .
( b ) the electric field at the oxygen site of the inner ( black ) and outer ( red ) layers as a function of the buckling distance . ] the doping of individual layers will be driven by madelung energy differences between the inner and outer planes . in order to determine the madelung potential @xmath62 of each site we again employ the ewald summation technique using the structural data from ref . and assigning formal valence charges to each atom . in principle , both the inner and outer cu sites , as well as the inner and outer o sites , can have different madelung energies . the in - plane difference @xmath63 , where @xmath64 is a plane index , determines the charge transfer energy@xcite @xmath65 for each plane .
the difference between layers @xmath66 , where @xmath67 cu or o , controls the relative site energies of the two planes .
however , our ewald calculations show that @xmath68 and @xmath69 .
we therefore neglect the differences and take the charge transfer energy to be the same in each layer but shift the outer plane s site energies by @xmath70 where @xmath71 is the dielectric constant which taken to be @xmath72 . to understand the self - doping phenomena we consider a simple tight - binding model for the cuo@xmath0 planes .
for simplicity , the form for the bandstructure is taken from a 5-parameter tightbinding fit to the low - energy dispersion of bi2212@xcite as a representative singleband model .
we also introduce the usual inter - planar coupling term@xcite @xmath73 with @xmath74 mev .
the resulting model hamiltonian is @xmath75\end{aligned}\ ] ] where @xmath76 is the plane index , @xmath77 for the outer planes and @xmath78 for the inner planes and @xmath79 is a sum over neighboring planes .
( the in - plane charge transfer energy has been absorbed into the definition of @xmath80 . ) finally , @xmath81 is the chemical potential which is adjusted to maintain the total filling of the parent compound , @xmath82 .
the resulting model is then diagonalized for a given @xmath83 in order to obtain the relative filling of the four planes . in previous lda treatments of this material@xcite , it was found that the identification of the carrier types in each layer is dependent on the degree of cu - o buckling occurring in the outer layers .
we therefore calculate the madelung potential difference between the inner and outer layers as a function of the degree of buckling .
the results are shown in fig . [
fig : madelung]a .
for flat or dimpled ( @xmath84 ) planes , @xmath85 is positive indicating a larger site energy in the outer layer .
charge is therefore expected to flow from the outer planes ( op ) to the inner planes ( ip ) and the op / ip are identified as @xmath18- and @xmath17-type respectively . as the plane is buckled outward @xmath83 is suppressed and becomes negative for @xmath86 @xmath87 at which point the identity of the planes is reversed .
the filling of the ip and op , obtained from diagonalizing ( [ eq : hamiltonian ] ) , is shown in the inset of fig .
[ fig : madelung]b
. once the fillings have been obtained we uniformly transfer charge between the planes and recalculate the crystal field at the oxygen sites .
the resulting fields are plotted in fig .
[ fig : madelung]b . for
a dimpled plane the electric field at the ip and op roughly follow a linear dependence on the buckling distance with the ip having only a slightly weaker field strength . in this case , the kink strength measured by arpes in the two bands would be comparable . for a buckled plane ,
once the sign of the madelung potential difference changes and the identification of the planes reverses , deviations in the field strength become more pronounced .
since the coupling enters as @xmath88 the @xmath17-type outer plane is expected to have a stronger kink feature , consistent with photoemission experiments .
the carrier concentration of the two sets of planes , determined from the luttinger fraction , were reported in ref . with dopings of @xmath89 and @xmath90 in the @xmath18- and @xmath17-type bands .
this observation , coupled with the different kink strength in each of the bands corresponds to @xmath91 @xmath87 in our calculations . for this degree of buckling ,
the ratio of the crystal fields is @xmath92 and the ratio of the kink strength is @xmath93 .
( the absolute value of @xmath94 depends on the form of the coupling constant @xmath95@xcite . )
this ratio is slightly larger than the ratio of @xmath96 reported in ref .
however the agreement is good considering the simplicity of the model .
it is also clear that the inequivalent dopings can be understood from electrostatic configurations .
furthermore , the asymmetry in the coupling in each layer falls naturally into the @xmath5-axis phonon scenario with the strength of the coupling being further driven by the self - doping process .
in this work we have presented aspects of the material and doping dependence of the dispersion renormalizations in the nodal and antinodal regions of various single- and multi - layer cuprates .
we have found that the strength of the nodal kink has a strong material dependence and varies with the number of layers present in the material . in general , the kink strength mirrors t@xmath2 , taking on a maximal value in the @xmath25 compounds .
the issue can be complicated further in the multi - layer cuprates , where madelung potential differences can lead to inequivalant dopings in the various layers .
this can lead to further symmetry breaking across the cuo@xmath0 planes , and results in different kink strength in the different layers within the same material . using a simple tight - binding model and electrostatic calculations we have developed a picture of this phenomena in self - doped f0234 , which is consistent with recent arpes studies .
the renormalization in the antinodal region also shows a marked dependence on the number of layers present in the material and is unresolved in the single layer cuprates .
this result is difficult to reconcile for coupling to the spin resonance mode , which is expected not to vary with the number of layers , but is naturally explained by coupling to the b@xmath6 phonon when one considers the crystal structure of these materials .
further evidence for multiple phonon modes was found in the doping dependence of bi-2212 . here
, the features in the nodal and antinodal regions exhibit different behavior . once gap referenced , the energy scale in the nodal region changes from @xmath97 mev to @xmath98 mev as the sample is overdoped .
this change of energy scales can not be explained by coupling to a single mode and therefore rules out the spin resonance mode , at least as the sole player . in the phonon scenario
this signifies a trade - off between dominant coupling to the b@xmath6 mode near optimal doping and a dominant coupling to the bond stretching mode in the overdoped samples .
the change of relative coupling is due to increased screening of the b@xmath6 mode as the carrier concentration is increased .
both the doping and materials dependence presented here provides compelling evidence that a spectrum of phonon modes are responsible for both the nodal and antionodal low - energy renormalizations observed in the cuprates .
this work was supported by the us department of energy , office of basic energy science under contracts de - ac02 - 76sf00515 and de - ac02 - 05ch11231 .
portions of this research were carried out at the stanford synchrotron radiation lightsource ( ssrl ) , a national user facility operated by stanford university on behalf of the us department of energy , office of basic energy sciences .
this research used resources of the national energy research scientific computing center , which is supported by the office of science of the u.s .
department of energy under contract no .
de - ac02 - 05ch11231 .
s. j. would like to acknowledge support from nserc and sharcnet .
99 p. v. bogdanov , a. lanzara , s. a. kellar , x. j. zhou , e. d. lu , w. j. zheng , g. gu , j .-
shimoyama , k. kishio , h. ikeda , r. yoshizaki , z. hussain and z .- x .
shen , phys .
lett . * 85 * , 2581 ( 2000 ) .
p. d. johnson , t. valla , a. v. fedorov , z. yusof , b. o. wells , q. li , a. r. moodenbaugh , g. d. gu , n. koshizuka , c. kendziora , sha jian , d. g. hinks , phys .
* 87 * , 177007 ( 2001 ) .
a. lanzara , p. v. bogdanov , x. j. zhou , s. a. kellar , d. l. feng , e. d. lu , t. yoshida , h. eisaki , a. fujimori , k. kishio , j .-
shimoyama , t. noda , s. uchida , z. hussain and z .- x .
shen , nature ( london ) * 412 * , 510 ( 2001 ) .
a. kaminski , m. randeria , j. c. campuzano , m. r. norman , h. fretwell , j. mesot , t. sato , t. takahashi and k. kadowaki , phys .
lett . * 86 * , 1070 ( 2001 ) .
t. cuk , d. h. lu , x. j. zhou , z .-
x.shen , t. p. devereaux and n. nagaosa , phys .
( b ) * 242 * 11 ( 2005 ) .
d. s. dessau , b. o. wells , z .- x .
shen , w. e. spicer , a. j. arko , r. s. list , d. b. mitzi and a. kapitulnik , phys .
lett . * 66 * , 2160 ( 1991 ) .
t. cuk , f. baumberger , d. h. lu , n. ingle , x. j. zhou , h. eisaki , n. kaneko , z. hussain , t. p. devereaux , n. nagaosa and z .- x .
shen , phys .
lett . * 93 * , 117003 ( 2004 ) .
a. d. gromko , a. v. fedorov , y .- d .
chuang , j. d. koralek , y. aiura , y. yamaguchi , k. oka , y , ando and d. s. dessau , phys .
b * 68 * , 174520 ( 2003 ) .
t. k. kim , a. a. kordyuk , s. v. borisenko , a. koitzsch , m. knupfer , h. berger and j. fink , phys .
lett . * 91 * , 167002 ( 2003 ) .
t. sato , h. matsui , t. takahashi , h. ding , h .- b .
yang , s .- c .
wang , t. fujii , t. watanabe , a. matsuda , t. terashima and k. kadowaki , phys .
rev . lett . * 91 * , 157003 ( 2003 ) .
m. r. norman , h. ding , j. c. campuzano , t. takeuchi , m. randeria , t. yokoya , t. takahashi , t. mochiku and k. kadowaki , phys .
lett . * 79 * , 3506 ( 1997 ) .
s. v. borisenko , a. a. kordyuk , v. zabolotnyy , j. geck , d. inosov , a. koitzsch , j. fink , m. knupfer , b. bchner , v. hinkov , c. t. lin , b. keimer , t. wolf , s. g. chiuzbain , l. patthey and r. follath , phys .
96 * , 117004 ( 2006 ) ; s. v. borisenko , a. a. kordyuk , a. koitzsch , j. fink , j. geck , v. zabolotnyy , m. knupfer , b. bcher , h. berger , m. falub , m. shi , j. krempasky and l. patthey , phys . rev
. lett . * 96 * , 067001 ( 2006 ) .
a. a. kordyuk , s. v. borisenko , v. b. zabolotnyy , j. geck , m. knupfer , j. fink , b. bcher , c. t. lin , b. keimer , h. berger , a. v. pan , s. komiya and y. ando , phys .
lett . * 97 * , 017002 ( 2006 ) .
w. meevasana , n. j. c. ingle , d. h. lu , j. r. shi , f. baumberger , k. m. shen , w. s. lee , t. cuk , h. eisaki , t. p. devereaux , n. nagaosa , j. zaanen and z .- x .
shen , phys .
lett . * 96 * , 157003 ( 2006 ) .
x. j. zhou , j. shi , t. yoshida , t. cuk , w. l. yang , v. brouet , j , nakamura , n. mannella , s. komiya , y. ando , f. zhou , w. x. ti , j. w. xiong , z. x. zhao , t. sasagawa , t. kakeshita , h. eisaki , s. uchida , a. fujimori , z. zhang , e. w plummer , r. b. laughlin , z. hussain and z .- x .
shen , phys .
lett . * 95 * , 117001 ( 2005 ) .
t. dahm , v. hinkov , s. v. borisenko , a. a. kordyuk , v. b. zabolotnyy , j. fink , b. bchner , d. j. scalapino , w. hanke and b. keimer , nature physics * 5 * , 217 ( 2009 ) .
p. w. anderson , science * 316 * , 1705 ( 2007 ) .
abanov , a. v. chubukov , m. eschrig , m. r. norman and j. schmalian , phys .
* 17 * , 177002 ( 2002 ) .
kee , s. a. kivelson and g. aeppli , phys .
lett . * 88 * , 257002 ( 2002 ) .
t. p. devereaux , t. cuk , z .- x .
shen and n. nagaosa , phys .
lett . * 93 * , 117004 ( 2004 ) .
w. s. lee , w. meevasana , s. johnston , d. h. lu , i. m. vishik , r. g. moore , h. hisaki , n. kaneko , t. p. devereaux and z. x. shen , phys .
b * 77 * , 140504(r ) ( 2008 ) .
w. meevasana , t. p. devereaux , n. nagaosa , z .- x .
shen and j. zaanen , phys .
b * 74 * , 174524 ( 2006 )
. h. isasawa , j. f. douglas , k. sato , t. masui , y. yoshida , z. sun , h. eisaki , h bando , a. ino , m. arita , k. shimada , n. namatame , m. taniguchi , s. tajima , s. uchida , t. saitoh , d. s. dessau and y aiura , phys .
* 101 * , 157005 ( 2008 ) .
j. lee , k , kujita , k. mcelroy , j. a. slezak , m. wong , t. aiura , h. bando , m. ishikado , t. masui , j .- x .
zhu , a. v. balatsky , h. eisaki , s. uchida and j. c. davis , nature ( london ) * 442 * , 546 ( 2006 ) .
n. jenkins , y. fansano , c. berthod , i. maggio - aprile , a. piriou , e. giannini , b. w. hoogenboom , c. hess , t. cren and .
fischer , arxiv:0909.0500v1 ( 2009 ) .
x . zhu , k. mcelroy , j. lee , t. p. devereaux , q. si , j. c. davis and a. v balatsky , phys .
lett . * 97 * , 177001 ( 2006 ) . j .-
zhu , a. v. balatsky , t. p. devereaux , q. si , j. lee , k. mcelroy and j. c. davis , phys .
b * 73 * , 014511 ( 2006 ) . y. chen , a. iyo , w. yang , a. ino , s. johnston , h. eisaki , h. namatame , m. taniguchi , t. p. devereaux , z. hussain and z .- x .
shen , phys .
lett . * 103 * , 036403 ( 2009 ) .
t. p. devereaux , a. virosztek , a. zawadowski , m. opel , p. f. mller , c. hoffmann , r. philipp , r. nemetschek , r. hackl , a. erb , e. walker , h. berger and l. forr , solid state comm . * 108 * , 407 ( 1998 ) ; m. opel , r. hackl , t. p. devereaux , a. virosztek , a. zawadowski , a. erb , e. walker , h. berger and l. forr , phys .
b * 60 * , 9836 ( 1999 ) .
s. johnston , f. vernay and t. p. devereaux , euro .
. lett . * 86 * , 37007 ( 2009 ) .
q. huang , j. w. lynn , q. xiong and c. w. chu , phys .
b * 52 * , 462 ( 1995 ) ; p. g. radaelli , j. l. wagner , b. a. hunter , m. a. beno , g. s. knapp , j. d. jorgensen and d. g. hinks , physica c * 216 * , 29 ( 1993 ) ; a. r. armstrong , w. i. f. david , i. gameson , p. p. edwards , j. j. capponi , p. bordet and m. marezio , phys . rev .
b * 52 * , 15551 ( 1995 ) ; m. paranthaman and b. c. chakoumakos , jour . of sol . state chem .
* 122 * , 221 ( 1996 ) ; q. huang , o. chmaissem , j. j. capponi , c. chaillout , m. marezio , j. l. tholence and a. santoro , physica c * 227 * , 1 ( 1988 ) .
p. ewald , ann .
369 * , 253 ( 1921 ) .
s. johnston , f. vernay , b. moritz , t. p. devereaux , z .- x .
shen , n. nagaosa and j. zaanen , to be published .
o. k. andersen , s. y savrasov , o. jepsen and a. i. liechtenstein , j. low temp .
phys . * 105 * , 285 ( 1996 ) .
w. s. lee , k. tanaka , i. m. vishik , d. h. lu , r. g. moore , h. eisaki , a. iyo , t. p. devereaux and z. x. shen , phys .
* 103 * , 067003 ( 2009 ) .
w. s. lee _
_ , in band renormalization effect of bi2223 , edited by a. bussmann - holder and h. keller ,
high t@xmath2 superconductors and related transition metal oxides " .
( springer - verlag , berlin heidelberg , 2007 ) .
j. wei , y. zhang , h. w. ou , b. p. xie , d. w. shen , j. f. zhao , l. x. yang , m. arita , k. shimada , h. namatame , m. taniguchi , y. yoshida , h. eisaki and d. l. feng , phys .
* 101 * , 097005 ( 2008 ) .
the peak - dip - hump structure in the single - layer bi-2201 is rather weak and appears abruptly in a narrow range of momentum space with a different energy scale than that along the nodal direction@xcite .
in addition , the issue of a peak- dip - hump structure in the single - layer compound seems more uncertain , as the reported peak - dip - hump lineshape appears to be inconsistent among the literature @xcite .
j. graf , m. dastuto , c. jozwiak , d. r. garcia , n. l. saini , m. krisch , k. ikeuchi , a. q. r. baron , h. eisaki and a. lanzara , phys .
lett . * 100 * , 227002 ( 2008 ) .
h. he , p. bourges , y. sidis , c. ulrich , l. p. regnault , s paihs , n. s. berzigiarova , n. n. kolesnikov and b. keimer , science * 295 * , 1045 ( 2002 ) .
t. p. devereaux , a. virosztek and a. zawadowski , phys .
b * 51 * , 505 ( 1995 ) ; t. p. devereaux , a. virosztek and a. zawadowski , phys , rev .
b * 59 * , 14618 ( 1999 ) . a. iyo , m. hirai , k. tokiwa , t. watanabe and y. tanaka , supercond .
* 17 * , 143 ( 2004 ) .
a. w. sandvik , d. j. scalapino and n. e. bickers , phys . rev .
b * 69 * , 94523 ( 2004 ) .
the nominal doping level is estimated from t@xmath2 using the empirical law @xmath99 $ ] .
j. meng , w. zhang , g. liu , l. zhao , h. liu , z. jia , w. lu , z. dong , g. wang , h. zhang , y. zhou , y. zhu , x. wang , z. zhao , z. xu , c. chen and x. j. zhou , phys .
b * 79 * , 024514 ( 2009 ) .
y. chen , a. iyo , w. yang , x. zhou , d. lu , h eisaki , t. p. devereaux , z. hussain and z .- x .
shen , phys .
97 * , 236401 ( 2006 ) .
m. r. norman , m. randeria , h. ding , j. c. campuzano , phys .
b * 52 * , 615 ( 1995 )
. t. xiang and j. m. wealthy , phys .
lett . * 77 * , 4632 ( 1996 ) .
w. xie , o. jepsen , o. k. andersen , y. chen and z .- x .
shen , phys .
lett . * 98 * , 047001 ( 2007 ) .
w. s. lee , s. johnston , t. p. devereaux and z .- x .
shen , phys .
b * 75 * , 195116 ( 2007 ) . | in this paper we present a review of bosonic renormalization effects on electronic carriers observed from angle - resolved photoemission spectra in the cuprates .
we specifically discuss the viewpoint that these renormalizations represent coupling of the electrons to the lattice , and review how the wide range of materials dependence , such as the number of cuo@xmath0 layers , and the doping dependence can be straightforwardly understood as arising due to novel electron - phonon coupling . |
accurate knowledge of thermonuclear reaction rates is important @xmath4 in understanding the generation of energy , the luminosity of neutrinos , and the synthesis of elements in stars . due to the coulomb barrier ( height @xmath5 ) of the entrance channel ,
the reaction cross section @xmath6 drops nearly exponentially with decreasing energy @xmath7 .
thus it becomes increasingly difficult to measure @xmath6 and to deduce the astrophysical @xmath8 factor defined by the equation @xmath9
@xmath10 with the sommerfeld parameter given by @xmath11 .
the quantities @xmath12 and @xmath13 are the nuclear charges of the interacting particles in the entrance channel , @xmath14 is the reduced mass ( in units of amu ) , and @xmath7 is the center - of - mass energy ( in units of kev ) . although experimental techniques have improved @xmath9 significantly over the years to extend @xmath6 measurements to lower energies , it has not yet been possible to measure @xmath6 within the thermal energy region in stars .
this region is determined by the gamow energy window @xmath15 ( the gamow peak ) for a given stellar temperature and lies far below the height of the coulomb barrier , approximately at @xmath16 . instead , the observed @xmath6 data at higher energies had to be extrapolated to thermal energies .
as always in physics , such an extrapolation into the unknown can lead to considerable uncertainties .
the low - energy studies of thermonuclear reactions in a laboratory at the earth s surface are hampered predominantly by the effects of cosmic rays in the detectors .
passive shielding around the detectors provides a reduction of gammas and neutrons from the environment , but it produces at the same time an increase of gammas and neutrons due to the cosmic - ray interactions in the shield itself .
a 4@xmath17 active shielding can only partially reduce the problem of cosmic - ray background .
an excellent solution is to install an accelerator facility in a laboratory deep underground @xcite . as a pilot project ,
a 50 kv accelerator facility has been installed @xmath18 in the laboratori nazionali del gran sasso ( lngs ) , where the flux of cosmic - ray muons is reduced by a factor 10@xmath19 @xcite .
the luna pilot project ( luna = laboratory for underground nuclear astrophysics ) was designed primarily for a renewed study of the @xmath0he(@xmath0he,2p)@xmath1hereaction ( @xmath20 = 12.86 mev ) in the energy range of the solar gamow peak ( @xmath15 = @xmath21 kev ) for a central star temperature of @xmath22 k. the reaction is a member of the hydrogen burning proton proton ( pp ) chain @xmath9 , which is predominantly responsible for the energy generation and neutrino luminosity @xmath23 of the sun .
so far , the reaction has been studied down to about 25 kev ( section ii ) but there remains the possibility of a narrow resonance at lower energies .
the hypothesis of a low energy resonance was first advanced @xmath24 as a solution to the solar neutrino puzzle , which in those days was regarded as a deficit of @xmath25b neutrinos .
for this purpose a resonance with @xmath26 21 kev and @xmath27 6 kev was considered @xmath28 as the most favourable case .
experimental data available at that time were not inconsistent with the existence of a resonance with @xmath29=15 - 20 kev and @xmath27 2 kev @xmath30 .
as more data on solar neutrinos became available , it became clear that the deficit of @xmath31be neutrinos is stronger than that of @xmath25b neutrinos .
it was shown @xmath30 that such a pattern of suppression occurs if @xmath26 21 kev .
such a resonance level in @xmath32be has been sought @xmath9 without success by various indirect routes , and it is also not predicted by most nuclear - structure theories .
however , the existence of this hypothetical resonance can be positively dismissed only by direct measurements at the required low energies ( i.e. within the solar gamow peak ) . for nuclear reactions studied in the laboratory ,
the target nuclei and the projectiles are usually in the form of neutral atoms / molecules and ions , respectively .
the electron clouds surrounding the interacting nuclides act as a screening potential : the projectile effectively sees a reduced coulomb barrier .
this in turn leads to a higher cross section , @xmath33 , than would be the case for bare nuclei , @xmath34 , with an exponential enhancement factor @xmath35 @xmath36 where @xmath37 is the electron - screening potential energy ( e.g. @xmath38 approximately , with @xmath39 an atomic radius ) .
it should be pointed out that for a stellar plasma the value of @xmath34 must be known because the screening in the plasma can be quite different from that in laboratory studies @xmath40 , and @xmath34 must be explicitly included in each situation .
thus , a good understanding of electron - screening effects is needed to arrive at reliable @xmath34 data at low energies .
low - energy studies of several fusion reactions involving light nuclides showed @xmath41 indeed the exponential enhancement of the cross section at low energies .
the observed enhancement ( i.e. the value of @xmath37 ) was in all cases close to or higher than the adiabatic limit derived from atomic - physics models .
an exception are the previous @xmath0he+@xmath0he data ( section ii ) , which show apparently a flat @xmath8 curve down to @xmath42 kev , although the effects of electron screening should have enhanced the data at 25 kev by about a factor 1.2 for the adiabatic limit @xmath37 = 240 ev .
thus , improved low - energy data are particularly desirable for this reaction .
we report here on such new data obtained by the luna collaboration within the solar gamow peak .
preliminary results , which have been published @xmath43 , are superseded by the present report .
the @xmath0he(@xmath0he,2p)@xmath1he reaction represents in the exit channel a 3-body - breakup : if the breakup is direct , one should observe a continuous energy distribution of the ejectiles described by phase - space considerations ; if the breakup follows a sequential process , the energies of the ejectiles are described by 2-body kinematics .
experiments have shown @xmath44 that at energies below @xmath45 mev the reaction proceeds predominantly via a direct mechanism and that the angular distributions approach isotropy with decreasing energy .
the @xmath8 energy dependence observed by various groups @xmath46 represents a consistent picture ( fig .
[ fig1 ] ) .
the only exception is the earliest experiment @xmath47 where @xmath8 is lower by a factor 3 to 5 compared to the other experiments ; the discrepancy is most likely caused by target problems ( @xmath0he trapped in an al foil ) . the absolute @xmath8 values of refs .
@xmath48 and @xmath49 ( as well as those from the present work , section vi ) agree - at overlapping energy regimes - within experimental uncertainties , while those of refs . @xmath50 and @xmath51 are lower by about 25% , suggesting a renormalization of their absolute scales .
however , in view of the relatively few data points and their relatively large uncertainties in comparison to the other data sets it has been suggested @xmath52 to omit these data , without significant loss of information .
we verified that no changes in the @xmath8 fit are appreciable ( within 1% ) by including or excluding the data sets from @xmath50 and @xmath51 .
a reaction mechanism was suggested @xmath53 at low energies , in which a neutron tunnels from one @xmath0he to the other , unimpeded by the coulomb barrier , up to a radial distance where the nuclei overlap appreciably . in this model ,
a diproton remains and subsequently fissions into 2 protons .
the calculated energy dependence of the @xmath8 factor described well the data ( dotted curve in fig .
[ fig1 ] ) , thus providing confidence in the extrapolation using a polynomial function ( solid curve in fig .
[ fig1 ] ) : @xmath54
technical details of the luna setup have been reported @xmath55 . briefly ,
the 50 kv accelerator facility ( fig . [ fig2 ] ) consisted of a duoplasmatron ion source , an extraction / acceleration system , a double - focusing 90@xmath56 analysing magnet ( with adjustable pole faces ) , a windowless gas - target system , and a beam calorimeter .
the energy spread of the ion source was less than 20 ev , the plasma potential energy deviated by less than 10 ev from the voltage applied to the anode , and the emittance of the source was 2 cm rad ev@xmath57 .
the ion source provided a stable beam current of about 1 ma over periods of up to 4 weeks .
the high voltage ( hv ) of the accelerator was provided by a power supply , which has a typical ripple of @xmath58 , a longterm stability of better than @xmath59 , and a temperature coefficient of better than @xmath60 /k .
the air - conditioned laboratory was kept at a temperature of @xmath61 c and a relative humidity of @xmath62% .
the hv of the accelerator was measured with a resistor chain , contained in an air - tight plexiglas tube , and a digital multimeter .
the resistor chain was built as a voltage divider , with fifty 20 m@xmath63 resistors and one 100 k@xmath63 resistor ( temperature coefficient = @xmath64/k ) .
the multimeter ( with a long term stability of @xmath65 per year ) provided the numerical value of the hv measured across the 100 k@xmath63 resistor .
this hv - measuring device was calibrated at the ptb in braunschweig ( germany ) at @xmath66c and @xmath67% to a precision of @xmath65 .
the beam entered the target chamber of the differentially pumped gas - target system ( 3 pumping stages ) through apertures of high gas flow impedance ( c to a in fig .
[ fig2 ] ) and was stopped in the beam calorimeter .
the gas pressure in the target chamber , @xmath68 , was measured with a baratron capacitance manometer to an accuracy of better than 1% .
this measurement was absolute and independent of the type of gas used . for @xmath0he gas ( 99.9% enriched in @xmath0he ) of @xmath68
= 0.50 mbar pressure , the system reduced the pressure to @xmath69 , @xmath70 , and @xmath64 mbar in the three upstream pumping stages ; a similar pressure reduction was observed for other @xmath68 values .
the gas composition in the target chamber was monitored with a mass spectrometer .
the @xmath0he gas was compressed by roots blowers , cleaned efficiently using a zeolite adsorption trap ( cooled to liquid nitrogen temperature ) , and fed back into the target chamber ( gas recirculation ) .
the pressure @xmath68 was kept at a constant value using a needle valve in combination with an electronic regulation unit .
as noted above , the main pressure drop occurred across the entrance aperture a ( 7 mm diameter , 40 mm length , 230 mm distance from aperture b ) .
it was shown that the pressure in the target chamber was essentially unmodified by the gas flow through the entrance aperture a ; thus , the geometrically extended target zone was characterized by a nearly static pressure .
beam - heating effects on the gas density are expected to be less than 0.5% @xmath71 for a maximum @xmath72 mbar @xmath0he target pressure and the 500 @xmath14a maximum @xmath0he@xmath73 beam current .
the beam current in the target area was determined to an accuracy of 3% using the beam calorimeter ( with a constant temperature gradient ) .
the calorimeter was placed ( fig .
[ fig3 ] ) at such a distance @xmath74 from the center of aperture a ( @xmath75 cm ) that angle straggling of the incident beam in the gas resulted in a beam profile smaller than the 200 mm@xmath76 active area of the calorimeter .
the luna facility was equipped with an interlock system , which allowed the system to run without an operator on site .
the duty time of the facility in the chosen running conditions was about 90% , with a weekly service time of 8 h. as the typical beam current in the target area was about 400 @xmath14a , a weekly charge of about 200 cb could be accumulated on the target .
the detection setup for the @xmath0he(@xmath0he,2p)@xmath1he studies had to fulfill the following requirements : 1 . a high absolute efficiency , in view of the expected reaction rates ( yields ) of about 1 event / day and less ; 2 .
a high rejection of natural radioactivity in the detectors , in the target chamber facing the detectors , and from the surrounding rocks at lngs ( mainly gammas ) ; tests at lngs have shown that commercial si detectors exhibited an intrinsic radioactivity level , which was about 200 times higher than the above reaction yield ; 3 . a high rejection of electronic noise , in view of the needed running times of several weeks per energy point ; 4 . a clear separation of the reaction products from those of @xmath0he(d , p)@xmath1he ( @xmath20 = 18.35 mev ) , due to deuterium contamination in the @xmath0hebeam ( as hd@xmath77 molecules of mass 3 ) and in the gas target ( found to be smaller ) .
this contaminant reaction has a cross section one millionfold higher than that of @xmath0he(@xmath0he,2p)@xmath1he at @xmath78 kev , mainly due to the barrier ratio @xmath5 ( d+ @xmath0he)/@xmath5 ( @xmath0he + @xmath0he ) = 0.56 , and thus extremely small deuterium contaminations ( of order 10@xmath79 ) can lead to sizeable event rates . in order to optimize the detection setup and to understand the resulting spectra for quantitative analyses , a monte carlo program @xmath80 was written to simulate the experiment under realistic conditions .
the monte carlo program produces energy and time spectra of the ejectiles as well as absolute yields , which could be compared directly with data .
various quantitative tests of the monte carlo predictions have been carried out successfully @xmath81 .
a monte carlo simulation @xmath80 of the proton spectrum from both reactions at a beam energy of 50 kev is shown in fig .
[ fig4 ] : a ni foil of @xmath0he and @xmath1he stopping thickness ( 20 @xmath14 m ) was placed in front of a 1000 @xmath14 m thick si detector ( maximum commercially available thickness at the time of the experimental project : the desirable thickness was 1400 @xmath14 m to stop the 14.7 mev protons from @xmath0he(d , p)@xmath1he ) ; a deuterium contamination of 10@xmath82 was assumed .
one sees the continuous energy distribution of the protons from @xmath0he(@xmath0he,2p)@xmath1he and the peak(s ) plus low - energy tail of protons from @xmath0he(d , p)@xmath1he ( due to incomplete stopping in the detector ) . both proton spectra overlap appreciably leading to large uncertainties in the yield analysis of the @xmath0he(@xmath0he,2p)@xmath1he events ; thus , the single spectra of si detectors did not represent a viable solution .
in addition , singles spectra would not reject the background due to natural radioactivity and electronic noise . by the requirement of proton - proton coincidences between detectors placed on opposite sides of the beam axis a unique signature of the @xmath0he(@xmath0he,2p)@xmath1he events could be obtained @xmath49 in the coincidence spectra ; also the background events ( radioactivity and electronic noise ) could essentially be eliminated . however , the price for this unique signature is a reduction of the absolute efficiency , by at least one order of magnitude compared to analysis via single spectra .
therefore , this approach was not followed ; instead the present approach utilized @xmath83-e detectors , in which single proton signals were detected in coincidence between the @xmath83 and e detectors of the telescope . in the detection setup ( fig . [ fig3 ] ) ,
designed according to the indications of monte carlo simulations , four @xmath84e - e telescopes ( placed in a rectangular target chamber ) were arranged around the beam axis : two opposite telescopes each at a distance of 2.7 cm from the beam axis and the two other telescopes each at 3.7 cm ; the distance from the center of the entrance aperture a to the center in front of the telescopes was @xmath85 cm .
the different distance from the center of the up - down and left - right couples of telescopes was due to mechanical constraints .
each telescope consisted of transmission surface barrier silicon detectors with a 0.25 @xmath14 m thick al layer deposited on both sides of the detectors .
the @xmath86e and e detectors both had an active square area of 2500 mm@xmath76 ; the @xmath86e ( and e ) detector had a thickness of 140 @xmath14 m ( and 1000 @xmath14 m ) and an energy resolution of 105 kev ( and 55 kev ) at @xmath87 = 5.5 mev . a mylar foil ( 1.2 @xmath14 m thick ) and an al foil ( 1.5 @xmath14 m thick ) were placed in front of each telescope ; they stopped the intense elastic scattering yield and shielded the detectors from beam induced light .
this double shielding was proved , in various background runs , to be effective and free from pinhole effects .
the detectors were maintained permanently at low temperature ( about -20@xmath56c ) using a liquid recirculating cooling system .
the @xmath1he ejectiles from @xmath0he(@xmath0he,2p)@xmath1he ( @xmath87 = 0 to 4.3 mev ) and @xmath0he(d , p)@xmath1he(@xmath87 = 3.7 mev ) were stopped in the @xmath86e detectors , while the ejected protons from @xmath0he(@xmath0he,2p)@xmath1he ( @xmath88 = 0 to 10.7 mev ) and @xmath0he(d , p)@xmath1he(@xmath88 = 14.9 mev ) left signals in both the @xmath86e and e detectors of a given telescope ( coincidence requirement ) .
[ fig5 ] shows the identification matrix of one @xmath86e - e telescope simulated with the monte carlo program at @xmath89= 50 kev and assuming a deuterium contamination of 10@xmath90 : the matrix reveals a clear separation of the events from both reactions .
standard nim electronics were used in connection with the four telescopes .
the signals were handled and stored using a camac multiparametric system , which allowed for on - line as well as for play - back data analyses .
signals from two or more detectors were considered coincident within a time window of 1 @xmath14s , making negligible the rate of random coincidences .
a pulser was permanently used in all detectors to check for dead time and electronic stability .
the acquisition system also stored concurrent information on experimental parameters ( such as ion beam current / charge , accelerator high voltage , and gas pressure in the target chamber ) via camac scalers .
controls have also been implemented to stop data acquisition if the beam was lost or to reject an event if an error in the data transmission occured .
the system handles safely a data flux of about 500 events / s . in the analysis of the data , the accepted events from @xmath0he(@xmath0he,2p)@xmath1he had to fulfill the following 3 conditions : 1 .
the events had to arise from proton - induced @xmath91e - e coincidences in a given telescope ; the coincidence requirement of each telescope essentially eliminated events due to natural radioactivity of the detectors themselves and of surrounding materials ; 2 .
the events had to lie in a closed region of the @xmath91e - e plane ( fig .
[ fig6 ] ) , whose borders were first deduced by monte carlo simulations and then fixed in order to cut the electronic noise ; 3 .
a proton event had to occur in only one given e - detector ; events which triggered more than one e - detector were rejected ; this condition reduced significantly events due to the residual electronic noise , but it also rejected some p - p coincidences .
the chosen @xmath91e - e energy region together with the anticoincidence requirement led to an absolute detection efficiency of @xmath92% as determined by the monte carlo program . in the @xmath91e - e region of events from @xmath0he(@xmath0he,2p)@xmath1he ( fig .
[ fig6 ] ) , no background events were ever observed at lngs since the installation of the equipment ( january 1994 ) during several no beam " and/or no target " background measurements ( up to two months running ) as well as during a 10 day run with a @xmath1he beam and a @xmath1he target gas . at the bochum laboratory ( earth surface )
a background rate of about 10 events / day was found with the same setup . with the discussed operating conditions the setup sensitivity in terms of cross section values is better than 10@xmath93 b.
although the selected regions in the spectra of the telescopes allow for a clear separation of the events from both reactions ( fig . [ fig6 ] ) , a few protons from the contaminant reaction @xmath0he(d , p)@xmath1hecan hit the detectors near the edges of their active volumes loosing only a fraction of their energy and thus leading to a background rate in the @xmath91e - e region of the @xmath0he(@xmath0he,2p)@xmath1he events .
the probability of such events was investigated by the monte carlo program as well as by direct measurements using projectiles with @xmath94 = 0.5 ( selected by the 90@xmath56 analysing magnet ) .
the ratio between the background events in the @xmath0he -@xmath0he region and those in the clearly separated d -@xmath0he region (= monitor ) turned out to be ( 0.40 @xmath95 0.04)% .
thus a deuterium contamination equal to 10@xmath79 allows @xmath96(e ) measurements as low as @xmath7 = 15 kev , with a signal - to - background ratio equal to 4 . during the reported experiments the deuterium contamination d/@xmath0he ranged between @xmath97 and @xmath98
for the isotropic emission of the protons in @xmath0he(@xmath0he,2p)@xmath1he ( section ii ) , the number of counts @xmath99 per unit of time arising from a differential length @xmath100 of the extended @xmath0he gas target is given by the expression @xmath101 where @xmath102 is the @xmath0he target density in units of atoms/@xmath103 ( derived from the measured target pressure @xmath68 , with a precision better than 1% ) , @xmath104 is the number of @xmath0he projectiles per unit of time ( derived from the beam calorimeter , with a precision of 3% ) , and @xmath105 is the absolute detection efficiency of all four telescopes including geometry and the acceptance criteria discussed in section iv .
the efficiency also takes into account that two protons are produced per reaction . introducing the stopping power @xmath106 ( i.e. the energy loss per unit length ) , equation ( 4 ) can be rewritten in the form @xmath107 the total number of counts for the full target length @xmath108 cm ( from the center of the aperture a to the beam calorimeter )
is then given by @xmath109 for the case of a thin target , i.e. introducing an effective beam energy @xmath110 corresponding to the mean value of the projectile energy distribution in the detection setup ( see below ) , one arrives at @xmath111 where the values for @xmath110 and the integral were derived from the monte carlo program . at subcoulomb energies a precise knowledge of the effective beam energy associated with observed yields is as important as the yield measurements themselves . for this reason , all monte carlo predictions
have been thoroughly tested @xmath80 . here
they are compared with a simple hand " calculation .
let us consider the case of an incident @xmath0heenergy of @xmath89 = 50.00 kev .
the absolute energy is known to a precision of @xmath95 2.5 ev , its long term stability is @xmath95 5.0 ev , and the gaussian energy spread ( fwhm ) is 20 ev ( section iii ) . when the beam passes through the gas target system filled with @xmath0he at the standard pressure of @xmath68 = 0.30 mbar , the beam loses energy in the 3 pumping stages and in the target chamber . for the calculation of this energy loss , we used stopping - power values given by the trim program @xmath112 .
it should be pointed out that experimental energy - loss data @xmath113 at the relevant low energies were found to be consistent with those from the trim program , within the experimental error of 10% .
we adopted this error in our analyses . with
@xmath114 ev / atom / cm@xmath76 at @xmath116 kev one finds an energy loss of @xmath117 ev in the 3 pumping stages and @xmath118 = ( 430 @xmath95 43 ) ev over the ( 8.3 @xmath95 0.1 ) cm distance from the center of aperture a to the center of the target chamber ( detector location ) : @xmath119 = ( 432 @xmath95 43 ) ev . the quoted error arises predominantly from the uncertainty in @xmath106 .
this leads to an effective energy at the center of the target chamber of @xmath120 kev .
the telescopes see an effective target length of @xmath121 cm ( see fig . [ fig8 ] ) .
the energy loss over this target length ( i.e. the target thickness ) is @xmath122 ev with an estimated error of @xmath95 36 ev . for a constant @xmath8 factor
the cross section drops by 11.2 % over this target thickness . assuming a linear decrease in cross section over this region @xmath9
, the effective energy evaluation gives @xmath123 kev , or a center - of - mass energy @xmath124 kev .
the error on @xmath110 transforms into a @xmath95 1.5% uncertainty on @xmath125 ) .
these estimates are consistent with monte carlo calculations simulating precisely the detection setup ( fig .
[ fig3 ] ) and the acceptance criteria of @xmath0he-@xmath0he events . the effective energy , as deduced from the mean energy of the distribution shown in fig .
[ fig7 ] , correspond to @xmath126 kev .
the spread of the energy distribution is determined by various effects : a ) the energy straggling @xmath127 at the center of the target , @xmath128 ev ; b ) the thermal doppler width @xmath9 at room temperature , @xmath129 ev ; c ) the energy spread over the target length seen by the detectors , @xmath130 ev . the total energy spread @xmath131 is therefore 171 ev , with an estimated error of @xmath132 ev .
this is in good agreement with the 189 ev spread predicted by the monte carlo program ( fig .
[ fig7 ] ) .
this procedure was applied for all beam energies and gas pressures .
it should be noted that the 10% systematic uncertainty in the projectiles energy loss in the target induces only a 0.1 0.2 % error on @xmath110 corresponding to a 1.0 3.5 % uncertainty in the @xmath8 values .
> from the geometry of the four square telescopes and the target length @xmath133 cm one expects an absolute efficiency over this length of the order of 10% ; the monte carlo simulations yield @xmath92% .
the geometrical setup efficiency was tested using an @xmath134-source , which was moved along the beam axis .
this test was possible , due to mechanical constraints , only recently with a new setup ( to be used in future work ) where eight 1000 @xmath14 m thick detectors were placed in two consecutive boxes ( each containing 4 detectors around the beam axis ) all at the same distance from the beam axis .
the results ( fig . [ fig8 ] ) are in good agreement with the corresponding monte carlo simulations ( solid curves in fig .
[ fig8 ] ) . to test the reliability of the corrections applied in the data analyses ,
we performed measurements at @xmath135 kev using different gas pressures : @xmath68 = 0.15 , 0.30 and 0.50 mbar ; for each pressure the target thickness , the energy loss , and the detection efficiency are different .
the resulting @xmath8 values ( table i ) are fully compatible with each other , well inside the experimental uncertainties . during another test period
the detection setup was changed : the detector box was moved 5.0 cm closer to the beam calorimeter extending the total target length from 32.2 cm to 37.2 cm .
the target pressure was fixed at @xmath68 = 0.50 mbar .
in this setup a significant decrease ( of the order of 25% ) in reaction yield was predicted by the monte carlo simulations compared to the standard setup and standard pressure ( @xmath68 = 0.30 mbar ) , due to the lower effective energy .
the @xmath8 results ( table 1 ) are within the experimental uncertainties in good agreement with values deduced for the standard setup .
table i and fig . [ fig9 ]
summarize the @xmath0he(@xmath0he,2p)@xmath1he results obtained until december 1996 with the luna setup at the 50 kv underground accelerator facility ( lngs ) .
the lowest counting rate was 3 events per day at @xmath136 20.76 kev .
the preliminary data @xmath80 have been completely re - analysed and have been integrated with the results of other new measurements .
the data obtained at higher energies ( 450 kv accelerator in bochum ) with the luna setup @xmath55 are also included for completeness .
previous literature data obtained at @xmath137 to 1080 kev @xmath138 are also shown in fig .
the luna data have been obtained at energies within the solar gamow peak , i.e. below the 22 kev center of this peak , and represent the first measurement of an important fusion cross section at energies near the center of the gamow peak .
no evidence of the hypothetical resonance can be found in the covered energy range . the observed energy dependence @xmath8 , for bare nuclides @xmath139 and shielded nuclides @xmath140 was parametrized using the expressions @xmath141 @xmath142 where @xmath143 , @xmath144 , @xmath145 , and @xmath37 are fit parameters .
the data set shown in fig .
[ fig9 ] was fitted , in the energy region between 20.7 and 1080 kev , using three methods : * first @xmath143 , @xmath144 , @xmath145 were obtained considering data for @xmath146 kev only ; then @xmath37 was deduced from all the data and the fixed @xmath139 expression ; * all the four parameters were allowed to change and the whole data set was fitted ; the resulting curves for @xmath139 and @xmath140 are shown in fig .
[ fig9 ] as dashed and solid curves , respectively ; * @xmath37 was fixed at 240 ev ( i.e. the value calculated in the adiabatic limit ) while @xmath143 , @xmath144 and @xmath145 were allowed to vary .
the results are summarized in table ii .
the three methods give compatible @xmath139 values while @xmath37 changes significantly : the methods a ) and b ) give @xmath37 values higher than the adiabatic limit ( 240 ev ) , consistent with observations in other fusion reactions .
it should be noted that fitting the luna data only with method b ) , a screening potential of 490@xmath147 30 ev is obtained .
the difference between observed and predicted @xmath37 values is not understood at present .
therefore we adopted @xmath3 from the theoretical adiabatic screening ( method c ) and derived a further error corresponding to the extreme cases @xmath148 ev ( @xmath149 mev b ) and @xmath150 ev ( @xmath151 mev b ) : @xmath152 where the first two errors arise from statistical and systematical uncertainties , respectively , and the last error from the lack of understanding of electron screening .
the difference between the luna data in the 20.7 - 24.8 kev region and @xmath8 calculated in the adiabatic limit ( i.e. fixing @xmath37=240 ev ) , can be used to estimate upper limits for the strength of a hypothetical resonance located at lower energies . for each measured point
we calculated an `` excess '' s - factor , @xmath153 from : @xmath154 where @xmath155(e ) are the data given in table i corrected for a 240 ev screening potential and @xmath156 comes from the fit c ) in table ii discussed above .
the excess cross section , integrated in the measured energy interval ( 20.7 - 24.8 kev ) , gives the integral yield : @xmath157 we may compare this value with the tail of a low - energy resonance integrated over the same energy range . for this purpose , we adopted the following procedure : the resonant cross section , @xmath158(e ) , was described according to the usual breit - wigner expression : @xmath159 where @xmath160 is the de broglie wavelength , @xmath161 is the statistical factor , @xmath162 is the resonance energy and @xmath163 is the resonance total width . due to the high energies of the protons in the exit channel , the partial width of the resonance decay , @xmath164 ,
was assumed to be energy - independent . for the partial width of the entrance channel , @xmath165
, we used the expression @xmath9 : @xmath166 where @xmath14 = 1.507 a.m.u .
is the reduced mass,@xmath167 = 4 fm is the nuclear radius and the dimensionless reduced width @xmath168 of the nuclear state has an upper limit of unity ( @xmath168 @xmath169 1 , wigner limit ) .
the penetrability @xmath170 is given by the equation : @xmath171 and was calculated for s - waves ( l=0 ) using the approximations for the coulomb wave functions , @xmath172 and @xmath173 , given in @xmath174 .
it turns out that @xmath175 1 kev at e @xmath169 25 kev ( for @xmath168 = 1 ) and thus @xmath176 .
equations ( 12),(13 ) and ( 14 ) lead to the expression : @xmath177 where the constant a contains the parameters @xmath14 , @xmath167 and @xmath168 . its value can be determined by the comparison with the experimental value @xmath178 : @xmath179 with this functional relation @xmath168 is known for fixed values of the resonance energy @xmath29 and of the total width @xmath180 .
it is then possible to evaluate the reaction rate @xmath181 of this hypothetical resonance and compare it with the non - resonant rate @xmath182 used in standard solar model ( ssm ) calculations .
[ fig10 ] shows the calculated ratio @xmath183 as a function of @xmath29 for some values of the total resonance width @xmath180 , assuming a central temperature of the sun of 15 @xmath184 k. the dotted curve in fig .
[ fig10 ] shows the reaction rate ratio in the wigner limit ( @xmath168 = 1 ) , which was calculated using equations ( 15 ) and ( 16 ) .
the @xmath181 values obtained with this procedure are upper limits for the reaction rate due to the hyphotetical resonance . at energies
@xmath185 9 kev one concludes that the presence of a resonance can not account for even a partial nuclear solution of the solar neutrino puzzle @xmath186 .
the same conclusion applies in the energy region between 9 and 20 kev for resonance widths @xmath187 2 kev .
room seems still available for a very narrow resonance in this interval where only direct measurements could rigorously dismiss ( or confirm ) its existence .
for the first time the @xmath0he(@xmath0he,2p)@xmath1he fusion reaction has been studied in an energy region deep inside the gamow peak , i.e. down to 20.7 kev .
this goal has been achieved at the new facility for nuclear astrophysics ( luna ) built in the underground laboratory of gran sasso ( italy).the results obtained so far show that the @xmath8 energy dependence is consistent with the predictions based on an extrapolation from higher energies . the presence of a low energy resonance in the @xmath0he(@xmath0he,2p)@xmath1he reaction , which could have strong effects on the solar neutrino problem " , does not emerge from the new data . in the near future
the luna collaboration will extend the measurements down to @xmath188 kev : the foreseen running time here is one year .
definite conclusions with respect to the expected solar neutrino fluxes have to await the results of these experiments . for this next phase a new detection set - up , designed to reduce
the background induced by deuterium contamination in the beam and in the gas target , has been developed and is now in a testing phase .
the electron screening effect , which produces an exponential enhancement of @xmath8 at low energies , can be observed in the new @xmath0he(@xmath0he,2p)@xmath1he data .
this is a second important result since previous data did not show a clear evidence of the enhancement ( fig .
[ fig1 ] ) .
the screening potential seems to be higher than the value predicted by the adiabatic model of the interaction between projectile and atomic electron clouds .
it should be noted that the adiabatic approximation , among other electron screening models , is that giving the highest screening potential .
a better quantitative determination of the screening potential will be possible when the luna collaboration will reach the lowest energy scheduled for the @xmath0he(@xmath0he,2p)@xmath1he experiment .
hopefully , a complete and clear picture of the screening effect will be possible when data collected at very low energies and for other fusion reactions are available .
this is one of the future goals of the luna collaboration .
the excess of the screening potential could also be explained as a tail of a narrow resonance lying in the not measured low - energy region .
upper limits for the strength of such a resonance have been calculated assuming the adiabatic limit for the screening potential .
unfortunately , these calculations can not exclude the existence of a narrow resonance ( width < 2 kev ) at energies between 9 and 20 kev with a strength high enough to give a sizeable contribution to the @xmath0he(@xmath0he,2p)@xmath1he reaction rate in our sun .
we are indebted with the director and all the staff of the laboratori nazionali del gran sasso for the hospitality and the support offered to this experiment w.a.fowler , rev.mod.phys .
56(1984)149 c.rolfs and w.s.rodney , cauldrons in the cosmos ( university of chicago press , 1988 ) g. fiorentini et al . , z.phys .
a350(1995)289 c. arpesella et al . , nuclear astrophysics at the gran sasso laboratory ( proposal of a pilot project with a 30 kv accelerator ) lngs 91/18 - december 1991 u.greife et al . ,
a350(1994)327 u.greife et al . , z.phys .
a351(1995)107 macro collaboration , phys .
b249(1990)149 j.n.bahcall and m.h.pinsonneault , rev.mod.phys .
64(1992)885 w.a.fowler , nature 238(1972)24 v.n.fetisov and y.s.kopysov , phys.lett .
b40(1972)602 v.n.fetisov and y.s.kopysov , nucl .
a239(1975)511 v. castellani , s .
deglinnocenti and g. fiorentini astron .
271(1993)601 h.j.assenbaum et al.,z.phys .
a327(1987)461 l. bracci et al . , nucl.phys
. a513(1990)316 b. ricci et al . , phys .
c52(1995)1095 p.prati et al . , z.phys .
a350(1994)171 k.langanke et al . , phys.lett b369(1996)211 c. arpesella et al . , phys.lett . b389(1996)452
a.d.bacher and t.a.tombrello , rev.mod.phys .
37(1965)433 m.r .
dwarakanath and h. winkler , phys.rev .
c4(1971)1532 a.krauss et al . , nucl.phys .
a467(1987)273 m.r .
dwarakanath , phys.rev .
c9(1974)805 n.m.wang et al . , sov.j.nucl.phys .
3(1966)777 w.m.good et al . , phys.rev .
83(1951)845 e.adelberger et al .
, workshop on solar fusion rates " , seattle , febr . 1997 ( to be submitted ) r.may and d.d.clayton , ap.j .
153(1968)855 j. goerres et al .
c.arpesella et al.,nucl.instr.meth .
a360(1995)607 j.f.ziegler , program trim-95 ( ibm research 1995 ) r.golser and d.semrad , nucl.instr.meth .
b69(1992)18 f.besenbacher et al . ,
40:9(1981)1 v. berezinsky , g .
fiorentini and m. lissia , physics letters b365(1996)185 - 192 m. abramowitz and j. a. stegun , handbook of mathematical functions , ( new york 1965 ) | we report on the results of the @xmath0he(@xmath0he,2p)@xmath1he experiment at the underground accelerator facility luna ( gran sasso ) . for the first time the lowest projectile energies utilized for the cross section measurement
correspond to energies below the center of the solar gamow peak ( @xmath2=22 kev ) .
the data provide no evidence for the existence of a hypothetical resonance in the energy range investigated .
although no extrapolation is needed anymore ( except for energies at the low - energy tail of the gamow peak ) , the data must be corrected for the effects of electron screening , clearly observed the first time for the @xmath0he(@xmath0he,2p)@xmath1hereaction .
the effects are however larger than expected and not understood , leading presently to the largest uncertainty on the quoted @xmath3 value for bare nuclides ( = 5.40 mev b ) . |
the @xmath1he nucleus is an example of a three - cluster system whose lowest threshold ( @xmath3 ) is a three - particle one .
such systems have a number of remarkable properties determined mainly by two factors the pauli principle and the character of the potential energy dependence due to the three - body structure .
the pauli principle imposes essential restrictions on the wave function of a system allowing it to contain only the components antisymmetric with respect to nucleon permutation .
thus , if we are using the expansion over the harmonic - oscillator basis , we must solve the problem of excluding the states forbidden by the pauli principle @xcite .
this , in particular , leads to the fact that the simplest wave function of the @xmath4-state of @xmath1he obtained with the translation - invariant shell model ( it coincides with the lowest oscillator - basis function ) is a superposition of states with three - particle hypermomentum @xmath5 ( @xmath6% ) and @xmath7 ( @xmath8 ) .
this function generates an infinite set of states differing by a number of oscillator quanta and having almost the same ratio of @xmath5 and @xmath7 components . the approximation using only this set of functions
can be appropriately called the @xmath0-approximation .
other basis states contain components with @xmath9 whose weight is small , as shown in @xcite .
notice that for @xmath10-states , wave functions of the @xmath0-approximation contain only @xmath7 components , while for the @xmath11-states they are the superpositions of @xmath12 ( @xmath13[1.ex]{\mbox{$\,\stackrel{\scriptstyle>}{\scriptscriptstyle \sim}\,$}}}}25\%}$ ] ) and @xmath14 ( @xmath13[1.ex]{\mbox{$\,\stackrel{\scriptstyle<}{\scriptscriptstyle \sim}\,$}}}}75\%}$ ] ) components .
the second characteristic feature of systems under consideration is the slow decrease of their potential energy for large values of the hyperradius @xmath15 .
as known ( see for example @xcite ) , in three - body systems , the potential energy has the asymptotical behaviour @xmath16 as @xmath17 , even if the two - body forces acting between any two of constituent particles are short - range .
this fact leads to two important results .
first , the boundary of a system is extremely loose and the rms radius is significantly larger compared with that of those neighbouring nuclei whose three - particle threshold is much higher than the binary one .
second , the phase shift of the three - body scattering rises sharply at low energies ( it is proportional to @xmath18 as @xmath19 ) which directly affects the behaviour of the electromagnetic transitions matrix elements between the ground state of @xmath1he and its continuous spectrum .
the aim of the present paper is to give a detailed quantitative analysis of general regularities in @xmath1he structure , based on a simple model taking into account the main features of a loosely - bound system with the lowest three - particle threshold .
the allowed basis functions for states with zero angular momentum @xmath20 of the @xmath0-approximation have the following form _
n(l=0 ) = a_0(n)_n0()u_0 ( ) + a_2(n)_n2()u_2 ( ) , where the superposition coefficients are a_0(n ) = , a_2(n ) = , @xmath21 are the hyperspherical functions , _ nk ( ) = n_nk ^k e^-^2/2 l^k+2_n(^2 ) are the normalized hyperradial harmonic - oscillator basis functions , @xmath22 are the laguerre polynomials , @xmath23 are the normalizing coefficients , @xmath24 therefore , the wave function of the @xmath25 state in the @xmath0-approximation is reduced to the expansion ( l=0 ) = _
[ 3 ] for all @xmath26 , the contribution of @xmath7 states is significantly greater than that of the @xmath5 states 0.86 <
a_2 ^ 2(n ) < 0.96 ; a_0 ^ 2(n ) = 1 - a_2 ^ 2(n ) , hence , in a simple model , the contribution of @xmath5 states can be neglected .
then , for the expansion ( [ 3 ] ) the following relations hold ( l=0 ) = _ n=0^ c_n _ n2()u_2 ( ) = ( ) u_2 ( ) = ( ) ^5/2 u_2 ( ) .
the function @xmath27 should satisfy the one - dimensional schrdinger equation + v()= e , [ 5 ] where @xmath28 is the nucleon mass , @xmath29 is the hypermomentum ( @xmath7 for @xmath25 ) , @xmath30 is the effective three - body potential .
the equation ( [ 5 ] ) is a starting point for the subsequent discussion .
it was solved by a numerical integration over @xmath15 from zero to a certain sufficiently large cutoff radius @xmath31 .
the effective potential is modelled by a function v ( ) = v_0 , [ 6 ] having correct asymptotics and without a singularity at @xmath32 . to reproduce the states with zero angular momentum
, we have chosen the parameters @xmath33 and @xmath34 such that the equation ( [ 5 ] ) would give experimental values for the @xmath1he binding energy and the rms radius , @xmath35 mev , @xmath36 fm .
the appropriate values of the parameters are v_0 = -87 mev , a = 3.073 fm . with these values
we have calculated the wave function of the @xmath1he ground state and @xmath4-states of its continuous spectrum .
the wave function of the ground state of @xmath1he together with the effective potential @xmath30 are presented in fig . 1
. the horizontal line below the @xmath15-axis corresponds to the ground state energy , its intersection with the potential energy curve marks the classical turning point .
the vertical line separates the values of @xmath15 less than @xmath37 fm .
the wave function falls as @xmath38 for small @xmath15 , due to the strong kinematical barrier , while for large @xmath15 , the long - range character of the potential @xmath30 leads to the slow decrease of the wave function and , respectively , to a significant diffuseness of the boundary of a nuclear system . as known
, loosely - bound binary systems with a short - range potential can be rather well approximated outside the potential range by the exponential function ( ) ( -r ) , = , [ 8 ] where @xmath39 is the bound state energy , @xmath40 is the radial variable of the binary channel .
the formal criterion of validity of such an approximation ( the zero - range approximation ) is the smallness of a ratio of the potential range and the cluster radii to the system radius @xmath41 expressed in terms of binding energy .
it seems useful to compare the exact wave function obtained after the solution of eq .
( [ 5 ] ) with the approximation given by eq . ( [ 8 ] ) ( for @xmath42 fm@xmath43 , dashed line in fig . 1 ) .
it can be seen that the nucleon system is much more loose than it is predicted by the wave function ( [ 8 ] ) .
the reason for that is not only that the function @xmath27 at small @xmath15 behaves as @xmath38 as @xmath44 , but first of all that the effective potential energy decreases slower than the exponential , while the formula ( [ 8 ] ) is obtained by supposing that the potential is short - range . as a result ,
the estimates of the square of hyperradius mean value @xmath45 based on ( [ 8 ] ) are about three times less than the experimental value .
of course , for sufficiently large @xmath15 the function @xmath27 has the asymptotics ( ) ck_4 ( ) c ( - ) , [ 10 ] where @xmath46 is the macdonald function , and the value of the coefficient @xmath47 ( it is related to the so - called nuclear vertex constants , see @xcite ) is considerably less than that of @xmath48 c 0.12 fm^-1/2 < 0.66 fm^-1/2 .
this asymptotics is denoted by short - dashed line in fig . 1 .
of the special interest are the solutions of eq .
( [ 5 ] ) for the continuous spectrum for relatively low over - threshold energies .
in particular , attention must be paid to the question of the behaviour of the three - to - three scattering phase @xmath50 as a function of energy in a potential field with the asymptotical behaviour @xmath51 and the powerful kinematic barrier @xmath52 corresponding to @xmath7 .
according to the estimates @xcite , at low energies , the phase shift , as is also demonstrated by our calculations , is proportional to @xmath53 ( or @xmath18 ) ak + ; a = 3.95 fm .
the low - energy values of @xmath54 calculated with different values of the cutoff radius @xmath31 are presented as functions of @xmath53 in fig .
2 . as seen , @xmath31 should be rather large , at least 1000 fm for energies below 10 kev .
the obtained value of @xmath55 gives a rather reasonable prediction for the ground - state energy of @xmath1he _ approx .
= ^22m1a^2 1.3 mev , ( cf .
= 0.97 mev ) .
the three - body phase shift @xmath50 calculated with different @xmath31 as functions of energy @xmath56 in an interval up to 5 mev are presented in fig .
as seen , for reliable calculations for medium - energy region ( @xmath57 mev ) @xmath31 may be taken about 50 fm and only for the lowest energies it must be increased .
the phase shift rises steeply from zero to values exceeding @xmath58 for @xmath59 and then slowly goes down . when the phase shift is near @xmath58 the first maximum of the wave function moves closer to zero so that the matrix element of the isoscalar transition from the ground state into the continuous spectrum increases ( see fig . 4 ) .
this matrix element reaches its maximal value for @xmath60 mev which could directly affect the electrodisintegration cross - section of @xmath1he .
the @xmath0-approximation satisfying the pauli principle for the @xmath10-states of @xmath1he contain only the hyperspherical function with @xmath7 .
therefore , to obtain the wave functions @xmath27 of @xmath10-states , we again turn to the equation ( [ 5 ] ) with @xmath7 but with the other parameters of the potential @xmath30 .
the @xmath1he nucleus has no @xmath10 bound state but it has a @xmath10-resonance at the energy of [email protected] mev with the width [email protected] mev @xcite . the energy and the width of the resonance can be reproduced with the following parameters of potential : v_0 = -92 mev , a = 2.834 fm . [
17 ] the wave function @xmath27 of the resonance state and the potential energy with the parameters ( [ 17 ] ) are presented in fig .
5 . for small @xmath15
the resonance wave function behaves similarly to the wave function of the ground state ( dashed line in fig . 5 )
but for larger @xmath15 , in the asymptotical region it , as it should be , oscillates .
the scattering phase shift in the @xmath10-state obtained by solving the equation ( [ 5 ] ) with the new potential parameters is presented in fig .
6 . at the energy of about 1
mev the phase shift has a typical behaviour for a resonance region .
we have also calculated the matrix element of the operator of the isoscalar @xmath62 transition from the ground state of @xmath1he to the @xmath10-states of its continuous spectrum .
the dependence of this matrix element upon the energy is presented in fig .
the narrow peak observed for the energy about @xmath63 mev again demonstrates the small width of @xmath10 resonance and the large value of the matrix element .
the simple three - cluster model based on the phenomenological long - range potential with the pauli principle taken into account has allowed us to reveal a number of regularities both for the weakly - coupled ground state of @xmath1he and for the states of its continuous spectrum for the relatively low ( up to a few mev ) energies .
the large value of the rms radius in the ground state can be explained by the considerable diffuseness of the wave function caused by the slowly - decreasing effective potential .
the asymptotic region , where the wave function decreases exponentially begins only at the hyperradius values greater than 1520 fm .
the phase shift of elastic scattering ( @xmath64 ) is proportional to @xmath53 or @xmath18 at low energies which leads to the sharp maximum of the matrix element of the isoscalar transition from the ground state to the @xmath4-states of the continuous spectrum .
finally , our calculations predict a considerable enhancing of the probability of the radiative capture of two neutrons by the alpha - particle for the energy corresponding to the @xmath10-resonance of @xmath1he . | by using the method of hyperspherical functions within the appropriate for this method @xmath0 approximation , the simple three - cluster model for description of the ground state and the continuous spectrum states of @xmath1he is developed .
it is shown that many properties of @xmath1he ( its large rms radius and large values of the matrix elements of electromagnetic transitions from the ground state into the continuous spectrum ) follow from the fact that the potential energy of @xmath1he system decreases very slowly ( as @xmath2 ) and the binding energy is small . |
in recent years the experimental study of neutrino oscillations has much contributed to our knowledge of particle physics by establishing non vanishing neutrino masses and by measuring or constraining the corresponding mixing angles . within the domain of neutrino oscillations , the main goal of the next generation of facilities
is the measurement of the mixing angle @xmath0 , which at present is only limited by an upper bound , and the observation of leptonic cp violation , for which we have no hints at the moment .
these measurements are extremely challenging , since the smallness of the present bound on @xmath4 at @xmath5 @xcite constrains the signals to be very subleading .
it is therefore necessary to have a good knowledge of the detector response function as well as an accurate knowledge of the beam to control the various systematic errors with a very good precision .
for this task , novel @xmath6 beams with extremely low backgrounds and systematic uncertainties have been proposed for future measurements based on the decay of muons ( neutrino factories @xcite ) or @xmath3 unstable ions ( @xmath3-beams @xcite ) . the @xmath7 fluxes from such decays would be extremely pure and the systematic errors very small , especially if the flux is normalized through a measurement with a near detector .
conversely , the systematic uncertainties associated with the detection process should be kept at the same accuracy level .
it is usually argued that the @xmath8 neutrino - nucleus cross section ratio can be understood at the level of some percent and that the @xmath7 cross section can be measured with similar precision at a near detector so that the @xmath9 signal can have systematic uncertainties of the order of a few percent at the far detector .
this , however , does not take into account our present ignorance of the actual value of the cross section at low energies and how different models lead to forecasted sensitivities to the unknown parameters that differ by much larger margins than the few percent uncertainty normally considered .
this question can severely impact the comparison of the relative performance of different facilities depending on the model adopted to parametrize their cross sections , particularly if the different facilities are sensitive to distinct energy regions . in this letter
we want to discuss more in detail this question showing that , when a realistic model of nuclear dynamics is adopted , the neutrino cross section can sizably affect the forecasted precision measurement of @xmath0 and the cp violating phase @xmath1 .
it is beyond the scope of this letter to discuss ( and critically revise ) all possible models for neutrino interactions , so we restrict ourselves to the relativistic fermi gas model ( rfg ) , variants of which are widely used in many montecarlo codes and , as examples of more refined calculations , to the one based on the spectral function approach ( sf ) , on the relativistic mean field ( rmf ) approximation and on the random phase approximation ( rpa ) , also including the effects of multinucleon contributions .
the main features of these models are briefly summarized in sect.[summary ] whereas in sect.[results ] we show how the measurement of @xmath0 and @xmath1 can be affected by the different descriptions of the nuclear dynamics .
we draw our conclusions in sect.[concl ] .
we work in the quasi - elastic regime ( qe ) , which is of interest in many current and planned experiments ( among them , miniboone has already released its first cross section measurement @xcite ) .
the doubly - differential cross section , in which a neutrino carrying initial four - momentum @xmath10 scatters off a nuclear target to a state of four - momentum @xmath11 can be written in born approximation as follows : [ elem ] = l _
w_a^ , where @xmath12 is the fermi constant and @xmath13 is the ckm matrix element coupling @xmath14 and @xmath15 quarks .
the leptonic tensor , that can be written in the form [ leptensor ] l_&=&8is completely determined by lepton kinematics , whereas the nuclear tensor @xmath16 , containing all the information on strong interactions dynamics , describes the response of the target nucleus .
its definition involves the initial and final hadronic states @xmath17 and @xmath18 , carrying four momenta @xmath19 and @xmath20 , respectively , as well as the nuclear electroweak current operator @xmath21 : [ hadronictensor ] w_a^&= & _ x 0 | j_a^^| x x | j_a^| 0 ^(4)(p_0 + q - p_x ) , where the sum includes all hadronic final states . the calculation of @xmath16 is a complicated task which deserves some approximation ; quite often the impulse approximation ( ia ) scheme is adopted , based on the assumptions that at large enough @xmath22 the target nucleus is seen by the probe as a collection of individual nucleons and that the particles produced at the interaction vertex and the recoiling @xmath23-nucleon system evolve independently . within this picture ,
the nuclear current can be written as a sum of one - body currents , i.e. @xmath24 , while the final state reduces to the direct product of the hadronic state produced at the weak vertex ( with momentum @xmath25 ) and that describing the @xmath23-nucleon residual system , with momentum @xmath26 : @xmath27 .
the calculation of the weak tensor as described in ref .
@xcite naturally leads to the concept of spectral function .
in fact , the final expression of the hadronic tensor can be cast in the following form : w_a^&= & d^3pde p(*p*,e ) w^(p , q ) , [ hadtensor ] where @xmath28 and the function @xmath29 is the target _ spectral function _
, i.e. the probability distribution of finding a nucleon with momentum @xmath30 and removal energy @xmath31 in the target nucleus .
it then encodes all the informations about the initial ( struck ) particle .
the quantity @xmath32 is the tensor describing the weak interactions of the @xmath33-th nucleon in free space ; the effect of nuclear binding of the struck nucleon is accounted for by the replacement @xmath34 with @xmath35 .
it follows that the second argument in the hadronic tensor is @xmath36 . substituting eq .
( [ hadtensor ] ) into eq .
( [ elem ] ) , we get the final formula for the _ nuclear _ cross section : [ final_cs ] = d^3pde p(*p*,e ) , in which we have redefined the _ elementary _ cross section as [ elem_cs ] = l _ w^ .
the calculation of @xmath29 has been only carried out for @xmath37 @xcite ; however , thanks to the simplifications associated with translation invariance , highly accurate results are also available for uniform nuclear matter , i.e. in the limit a @xmath38 with z = a/2 @xcite ( @xmath39 denotes the number of protons ) .
the spectral functions for medium - heavy nuclei have been modeled using the local density approximation ( lda ) @xcite , in which the experimental information obtained from nucleon knock - out measurements is combined with the results of theoretical calculations of the nuclear matter @xmath29 at different densities .
the rfg @xcite model , widely used in montecarlo simulations , provides the simplest form of the spectral function : [ fermigas ] p_rfgm(*p*,e)=()(p_f-*p * ) ( e_*p*-e_b+e ) , where @xmath40 is the fermi momentum and @xmath41 is the average binding energy , introduced to account for nuclear binding .
the term in parenthesis is a constant needed to normalize the spectral function to the number of target nucleons , @xmath42 .
thus , in this model @xmath40 and @xmath41 are two parameters that are _ adjusted _ to reproduce the experimental data . for oxygen , the analysis of electron scattering data yields @xmath43 mev and @xmath44 mev @xcite .
within the relativistic mean field approximation we refer to the model described in @xcite , where , like in the previous cases , the nuclear current is written as a sum of single - nucleon currents .
the wave functions for the target and the residual nuclei are described in terms of an independent - particle model .
then , the transition matrix elements can be cast in the following form : @xmath45 where @xmath46 and @xmath47 are the wave functions for initial bound and final outgoing nucleons , respectively , and @xmath48 is the relativistic current operator . in particular , the relativistic bound - state wave functions ( for both initial and outgoing nucleons ) are obtained as a solution of the dirac equation , in the presence of the same relativistic nuclear mean field potential , derived from a lagrangian containing @xmath49 , @xmath50 and @xmath51 mesons .
the calculated cross sections correctly account for the inclusive cross section which are interested in .
the last model we want to take into account has been introduced in @xcite , where the hadronic tensor is expressed in terms of the nuclear response functions treated in the random phase approximation ( rpa ) .
the response functions are related to the imaginary part of the corresponding full polarization propagators and the introduction of the rpa approximation means that the polarization propagators are the solutions of integral equations involving the bare propagators and the effective interaction between particle - hole excitations . within this formalism
, the authors of @xcite were able to show that multinucleon terms sizably increase the genuine charged current qe cross section @xcite in such a way to reproduce the miniboone results @xcite .
the mechanism responsible for the enhancement that brings the theoretical cross section into agreement with the data is multi nucleon knock out , leading to two particle - two hole ( 2p2h ) nuclear final states . in the following
, we will refer to this `` generalized '' qe cross section as rpa-2p2h whereas we adopt the short rpa for the genuine qe cross section . to summarize this section , we present in fig.[fig : xsect ] a comparison of the five total qe cross sections for the @xmath52 process ( left panel ) and @xmath53 ( right panel ) , in the energy range @xmath54 $ ] gev .
the curves have been computed using the dipole structure of the form factors and , in particular , a value of the axial mass close to @xmath55 gev . _ total charged current qe cross sections for the @xmath56 ( left panel ) and @xmath57 ( right panel ) processes in the energy range @xmath54 $ ] gev._,title="fig : " ] _ total charged current qe cross sections for the @xmath56 ( left panel ) and @xmath57 ( right panel ) processes in the energy range @xmath54 $ ] gev._,title="fig : " ] as it can be easily seen in the left panel , the rfg prediction sizably overestimates the sf , rmf and rpa results by roughly @xmath58 , a fact that is well known to happen also for many other models with a more accurate description of the nuclear dynamics than the rfg approach ( see , for instance , @xcite ) . on the other hand ,
the inclusion of 2p2h contributions largely enhances the qe cross section in the rpa approximation for energies above @xmath59 gev , although it is smaller than the rfg at smaller energies . for antineutrinos ( right panel )
the observed pattern is almost the same . from this comparison ,
a _ qualitative _ understanding of the impact of different models of neutrino - nucleus cross sections on the forecasted sensitivity of future facilities can be already derived
. a _ quantitative _ analysis of these effects will be the subject of the next section .
to estimate the impact of different models of the cross section on the measurement of @xmath0 and leptonic cp violation we choose a @xmath60 @xmath3-beam facility as a representative example .
the choice is motivated because the neutrino flux from such a facility spans up to @xmath61 gev with the peak around @xmath62 gev and is thus mostly sensitive to the quasielastic region explored here .
the @xmath3-beam concept was first introduced in ref .
@xcite and involves the production of @xmath3-unstable ions , accelerating them to some reference energy , and allowing them to decay in the straight section of a storage ring , resulting in a very intense and pure @xmath7 or @xmath63 beams .
the @xmath64 `` golden channel '' , which has been identified as the most sensitive to all the unknown parameters @xcite , can be probed at a far detector .
we have considered here the original @xmath3-beam proposal , where @xmath65ne ( @xmath66he ) ions are accelerated to @xmath67 at the cern sps and stored so that @xmath7 ( @xmath68 ) beams are produced and the golden channel oscillation is searched for at a mton class water cerenkov detector located at @xmath69 km at the frejus site , detailed analyses of the physics performance of this setup can be found in refs .
@xcite . in order to simulate the detector response when exposed to such a beam both in terms of signal efficiency and background , we have used the migration matrices derived in ref . @xcite .
systematic errors of @xmath70 and @xmath71 in the signal and background respectively have been taken into account .
in all the simulations the following best fit values and @xmath72 errors for the known oscillation parameters were assumed @xcite @xmath73 ev@xmath74 , @xmath75 ev@xmath74 , @xmath76 and @xmath77 .
these parameters were marginalized over to present the final curves .
the evaluation of the performance of the facility made use of the globes software @xcite .
it is important to notice that we are only using the quasielastic contribution to the neutrino cross section depicted in fig.[fig : xsect ] . as an illustration we have focused on the dependence on the nuclear model adopted of two different observables , namely the _ cp and @xmath0 discovery potentials _ , defined as the values of the cp - violating phase @xmath78 and @xmath0 for which respectively the hypothesis of cp conservation @xmath79 or @xmath80
can be excluded at 3@xmath49 after marginalizing over all other parameters .
the cp discovery potential is shown in fig.[fig : disc ] , where we superimposed the results obtained using the rfg cross section and the sf , rmf and rpa calculations .
_ left panel : cp discovery potential in the @xmath81 plane .
solid lines refer to the sf model , dotted lines to the rfg , short - dashed lines to the rmf , dot - dashed to rpa and long - dashed to rpa-2p2h .
right panel : the cp fraction._,title="fig : " ] _ left panel : cp discovery potential in the @xmath81 plane .
solid lines refer to the sf model , dotted lines to the rfg , short - dashed lines to the rmf , dot - dashed to rpa and long - dashed to rpa-2p2h .
right panel : the cp fraction._,title="fig : " ] in the left panel , the cp discovery potential is represented in the @xmath81-plane ; we clearly see that , for @xmath82 ( where the sensitivity is maximal ) the rfg model gives a prediction which is around a factor 2 better than the sf , rmf and rpa models ( which , as expected , behave almost in the same way ) in @xmath83 and around a @xmath84 better than the rpa-2p2h .
this is not surprising because the @xmath3-beam facility used in our simulations mainly probes energies smaller than @xmath85 gev , where the rfg is still larger than any other model ( see fig.[fig : xsect ] ) . for the other points in the parameter space , the difference is less evident but still significant .
the same information can be summarized in the right panel making use of the _ @xmath1 fraction _ ( @xmath86 ) , that represents the fraction of values of @xmath78 for which cp can be discovered at a given @xmath0 .
the @xmath1 fraction has a maximum around @xmath87 where the rfg model predicts @xmath88 and the other models give @xmath89 .
_ left panel : @xmath0 discovery potential in the @xmath81-plane .
solid lines refer to the sf model , dotted lines to the rfg , short - dashed lines to the rmf , dot - dashed to rpa and long - dashed to rpa-2p2h .
right panel : the cp fraction._,title="fig : " ] _ left panel : @xmath0 discovery potential in the @xmath81-plane .
solid lines refer to the sf model , dotted lines to the rfg , short - dashed lines to the rmf , dot - dashed to rpa and long - dashed to rpa-2p2h .
right panel : the cp fraction._,title="fig : " ] the difference in the sensitivity to @xmath0 can be seen in fig .
[ fig : disc2 ] where we show the results in the @xmath90-plane . in this case
the predictions of the sf , rmf and rpa models differ by up to a factor of @xmath91 compared to the rfg model for @xmath82 while the difference is less pronounced for @xmath92 .
the rpa-2p2h results are more similar to the rfg for @xmath92 and to the other models for @xmath93 .
this is also evident in the right panel where we show the cp fraction . finally , it is interesting to observe the effect of using different nuclear models also in the simultaneous determination of @xmath0 and @xmath78 .
we present an example in fig.[fig : patata ] where , for the input value @xmath94 , indicated with a dot , we show the capability of the @xmath3-beam to reconstruct the true values of our observables at @xmath95 cl . _
@xmath95 cl contour for the input value @xmath94 .
solid lines refer to the sf model , dotted lines to the rfg , short - dashed lines to the rmf , dot - dashed to rpa and long - dashed to rpa-2p2h .
_ ] for the sake of simplicity , we did not include degenerate solutions coming from our ignorance of the octant of @xmath96 and the hierarchy in the neutrino mass ordering but only the so - called `` intrinsic '' one .
the main feature here is that using the rfg and rpa-2p2h models we are able to reconstruct the true values of @xmath0 and @xmath78 within reasonable uncertainties , whereas with the other models we can only measure two distinct disconnected regions ( the fake one around the value of @xmath0 and @xmath93 ) , which worsen the global sensitivity on those parameters .
the effects we have mentioned have been generalized to account for the following cases : * the value of the axial mass does not fill the gap among the sf approach and the rfg model , at least in the range @xmath97 $ ] gev .
this is not surprising and has been already the subject of an extensive analysis @xcite ; * the same sensitivity behaviour as observed in figs.[fig : disc]-[fig : disc2 ] has been also seen for a different nuclear target , namely @xmath98fe .
this points to the conclusion that the rfg model overestimates the sensitivities to @xmath0 and @xmath78 for a vast class of nuclear targets .
in this letter we have analyzed the impact of different neutrino - nucleus charged current qe cross sections on the forecasted sensitivity of the future neutrino facilities to the parameters @xmath0 and @xmath78 .
we considered five different calculations , based on the rfg model ( widely used in the montecarlo codes ) and those based on the spectral function approach , on the relativistic mean field approximation and of the random phase approximation ( including multinucleon contributions , also ) in the quasi - elastic regime .
we found that the sensitivities computed from sf , rmf and rpa ( without the multinucleon contribution ) models are worse than the rfg results , for both @xmath0 and @xmath78 , by up to a factor 2 when an oxygen nuclear target is used to compute the event rates . to a less extent , this is also true for the rpa-2p2h model , due to the fact that the quasi - elastic cross section is larger than the genuine quasi - elastic cross section .
these variations with the nuclear models are so large that they can not be taken into account with the few percent systematic error uncertainty expected from dedicated measurements at near detectors .
indeed , the present uncertainty is much larger and allows for discrepancies between models that can strongly affect the forecasted sensitivity of a given facility , implying that special care must be adopted when comparing the performance of different facilities .
we have also checked that other nuclear targets produce quantitatively similar results .
this suggests that the use in montecarlo simulations of more refined nuclear models for the neutrino - nucleus interaction is mandatory , especially for those facilities whose bulk of events is in the quasi - elastic regime .
we are strongly indebted with omar benhar for providing us the oxygen and iron spectral functions . we also want to thank maria barbaro , juan antonio caballero and jose udias for providing us the neutrino - nucleus cross sections for the rmf approach , marco martini for the cross sections in the rpa and magda ericson and sergio palomares ruiz for useful comments on the manuscript .
d.m . was supported by the deutsche forschungsgemeinschaft , contract wi 2639/2 - 1 .
e.f.m was supported by the max - planck - institut fr physik , where this study was commenced , and from cern , where it was finalized .
e.f.m also acknowledges support from the european community under the european commission framework programme 7 design study : euronu , project number 212372 .
99 m. c. gonzalez - garcia and m. maltoni , phys .
* 460 * ( 2008 ) 1 [ arxiv:0704.1800 [ hep - ph ] ] . s. geer , phys . rev .
d * 57 * ( 1998 ) 6989 [ erratum - ibid .
d * 59 * ( 1999 ) 039903 [ arxiv : hep - ph/9712290 ] .
a. de rujula , m. b. gavela and p. hernandez , nucl .
b * 547 * ( 1999 ) 21 [ arxiv : hep - ph/9811390 ] .
p. zucchelli , phys .
b * 532 * ( 2002 ) 166 ; arxiv : hep - ex/0107006 .
a. a. aguilar - arevalo _ et al . _ [ miniboone collaboration ] , phys . rev .
d * 81 * ( 2010 ) 092005 [ arxiv:1002.2680 [ hep - ex ] ] .
o. benhar , n. farina , h. nakamura , m. sakuda and r. seki , phys .
d * 72 * , 053005 ( 2005 ) [ arxiv : hep - ph/0506116 ] ; o. benhar and d. meloni , nucl .
a * 789 * , 379 ( 2007 ) [ arxiv : hep - ph/0610403 ] .
dieperink , t. de forest and i. sick , phys . lett .
* b63 * , 261 ( 1976 ) ; c. ciofi degli atti , e. pace and g. salm , phys . rev .
c * 21 * , 805 ( 1980 ) ; h. meier - hajduk , ch .
hadjuk and p.u .
sauer , nucl .
* a395 * , 332 ( 1983 ) ; c. ciofi degli atti , s. liuti and s. simula , phys .
c * 41 * , r2474 ( 1990 ) ; h. morita and t. suzuki , prog .
phys . * 86 * , 671 ( 1991 ) ; o. benhar and v.r .
pandharipande , phys .
c * 47 * , 2218 ( 1993 ) .
o. benhar , a. fabrocini and s. fantoni , nucl . phys .
* a505 * , 267 ( 1989 ) ; a. ramos , a. polls and w.h .
dickhoff , nucl . phys . *
a503 * , 1 ( 1989 ) .
o. benhar , a. fabrocini , s. fantoni and i. sick , nucl . phys . *
a579 * , 493 ( 1994 ) .
r. a. smith and e. j. moniz , nucl .
b * 43 * ( 1972 ) 605 [ erratum - ibid .
b * 101 * ( 1975 ) 547 ] ; e. j. moniz , phys
. rev . * 184 * , 1154 ( 1969 ) .
j. m. udias , j. a. caballero , e. moya de guerra , j. r. vignote and a. escuderos , phys .
c * 64 * , 024614 ( 2001 ) [ arxiv : nucl - th/0101038 ] ; c. maieron , m. c. martinez , j. a. caballero and j. m. udias , phys .
c * 68 * , 048501 ( 2003 ) [ arxiv : nucl - th/0303075 ] ; j. a. caballero , j. e. amaro , m. b. barbaro , t. w. donnelly , c. maieron and j. m. udias , phys .
lett . * 95 * , 252502 ( 2005 ) [ arxiv : nucl - th/0504040 ] ; m. c. martinez , p. lava , n. jachowicz , j. ryckebusch , k. vantournhout and j. m. udias , phys .
c * 73 * , 024607 ( 2006 ) [ arxiv : nucl - th/0505008 ] .
o. benhar and d. meloni , phys .
d * 80 * , 073003 ( 2009 ) [ arxiv:0903.2329 [ hep - ph ] ] .
a. cervera , a. donini , m. b. gavela , j. j. gomez cadenas , p. hernandez , o. mena and s. rigolin , nucl .
b * 579 * ( 2000 ) 17 [ erratum - ibid .
b * 593 * ( 2001 ) 731 ] [ arxiv : hep - ph/0002108 ] .
m. mezzetto , j. phys .
g * 29 * ( 2003 ) 1771 [ arxiv : hep - ex/0302007 ] .
m. mezzetto , nucl .
* 143 * ( 2005 ) 309 [ arxiv : hep - ex/0410083 ] .
a. donini , e. fernandez - martinez , p. migliozzi , s. rigolin and l. scotto lavina , nucl .
b * 710 * ( 2005 ) 402 [ arxiv : hep - ph/0406132 ] .
a. donini , e. fernandez - martinez and s. rigolin , phys .
b * 621 * ( 2005 ) 276 [ arxiv : hep - ph/0411402 ] .
a. donini , d. meloni and s. rigolin , eur .
j. c * 45 * ( 2006 ) 73 [ arxiv : hep - ph/0506100 ] .
p. huber , m. lindner , m. rolinec and w. winter , phys .
d * 73 * ( 2006 ) 053002 [ arxiv : hep - ph/0506237 ] .
j. e. campagne , m. maltoni , m. mezzetto and t. schwetz , jhep * 0704 * ( 2007 ) 003 [ arxiv : hep - ph/0603172 ] . w. winter , phys .
d * 78 * ( 2008 ) 037101 [ arxiv:0804.4000 [ hep - ph ] ] .
w. winter , arxiv:0809.3890 [ hep - ph ] .
e. fernandez - martinez , nucl .
b * 833 * ( 2010 ) 96 [ arxiv:0912.3804 [ hep - ph ] ] .
j. burguet - castell , d. casper , e. couce , j. j. gomez - cadenas and p. hernandez , nucl .
b * 725 * , 306 ( 2005 ) [ arxiv : hep - ph/0503021 ] .
p. huber , m. lindner and w. winter , comput .
commun .
* 167 * ( 2005 ) 195 [ arxiv : hep - ph/0407333 ] .
p. huber , j. kopp , m. lindner , m. rolinec and w. winter , comput .
commun .
* 177 * ( 2007 ) 432 [ arxiv : hep - ph/0701187 ] . | we investigate how models for neutrino - nucleus cross sections based on different assumptions for the nuclear dynamics affect the forecasted sensitivities to neutrino oscillation parameters at future neutrino facilities .
we limit ourselves to the quasi - elastic regime , where the neutrino cross sections can be evaluated with less uncertainties , and discuss the sensitivity reach to @xmath0 and @xmath1 at a prototype low-@xmath2 @xmath3-beam , mostly sensitive to the quasi - elastic regime .
cern - ph - th/2010 - 225 + euronu - wp6 - 10 - 26 2.5 cm 0.2 cm 0.5 cm enrique fernandez martinez + max - planck - institut fr physik ( werner - heisenberg - institut ) , + fhringer ring 6 , 80805 mnchen , germany and cern , ph - th division , ch-1211 genve 23 , switzerland . + .1 cm .2 cm davide meloni + .1 cm institut fr theoretische physik und astrophysik , + universitt wrzburg , d-97074 wrzburg , germany + 0.7 cm 2truecm |
the observed properties and evolution of galaxies in the universe require the existence of mechanisms regulating and quenching star formation .
models and simulations which are not implementing these _ feedback _ processes predict e.g. galaxies with far larger stellar masses than are observed ( e.g. katz , weinberg & hernquist 1996 ; somerville & primack 1999 ; cole et al .
2000 ; springel & hernquist 2003 ; keres et al .
2009 , and references therein ) .
stellar winds , the explosion of supernovae and winds from active galactic nuclei ( agn ) all contribute to the heating and expulsion of gas from the disc and the bulge of galaxies into the halo and possibly far away into the lower density gas .
the signatures of strong winds are widely observed in local star forming galaxies ( see e.g. * ? ? ? * for a review ) .
their presence at high redshift has been probed spectroscopically by the observed offsets between the redshifts of the nebular emission lines and those of the inter - stellar absorption lines in spectra of lensed or extremely bright lyman - break galaxies ( lbgs , e.g. pettini et al .
2001 ; shapley et al . 2003
; martin 2005 ; rupke et al .
2005 ; tremonti et al . 2007 ; weiner et al . 2009 ) and has been explored in cosmological hydrodynamical simulations ( e.g. * ? ? ?
* ; * ? ? ?
recently , direct observational evidence for powerful , quasar - driven outflows has emerged ( e.g. feruglio et al .
2010 ; rupke & veilleux 2011 ; sturm et al . 2011 ; greene , zakamska & smith 2012 ; maiolino et al . 2012
; cicone et al . 2014 ; harrison et al . 2014 ) .
these outflows extend over scales of several kpc from the nucleus and have been detected in both atomic and molecular gas .
the indisputable evidence of the existence of galactic outflows leaves open one fundamental question : how far is this gas travelling ?
most of the times , the measured outflowing velocities are large enough to allow the escape of the gas from the virial radius ( e.g. pettini et al . 2001 ; vanzella et al .
2010 ) and theoretical models predict that the expanding material may reach characteristic distances of a few hundred kpc and up to @xmath13 mpc ( e.g. aguirre et al . 2001 ; calura & matteucci 2006 ; samui et al . 2008 ) .
the presence of products of the stellar nucleosynthesis outside galaxies has been confirmed by the first high - resolution , high signal - to - noise ratio ( snr ) optical spectra of high redshift quasars used as background bright sources . in those spectra ,
weak absorption features mainly due to the ionic doublet of at @xmath14 ( @xmath15 ) were detected at the same redshift of about half of the lyman @xmath16 ( ) lines with column density @xmath17 and of all lines with @xmath18 @xcite . in the same years , thanks to theoretical models and hydro - dynamical simulations , the so called forest of lines , observed in the spectra of high - redshift quasars , was interpreted as due mainly to the density fluctuations in the diffuse gas between galaxies : the inter - galactic medium ( igm , see e.g. cen et al .
1994 ; petitjean et al . 1995 ; or rauch 1998 ; weinberg et al .
1999 , for reviews ) .
the presence of heavy elements in gas outside the virial radius of galaxies confirmed observationally the effectiveness of outflows and generated a new question : have the metals , that we observe associated with the forest at @xmath19 , been ejected by the co - eval population of star forming galaxies or are they due ( also ) to previous generations of galaxies ?
indeed , models of the re - ionization process foresee that the same sources which re - ionized the universe were also able to pollute the gas to a metallicity floor of the order @xmath20 ( e.g. * ? ? ?
the volume filling factor of the metals due to this pre - enrichment depends on the nature of the sources , with the star - forming haloes with the smallest masses being the most efficient pollutant of the igm on large scales ( e.g. * ? ? ?
* ) . to test observationally this scenario ,
efforts have been concentrated on two main approaches : the detection of metals in the outskirts of galaxies and the investigation of the level of pollution in the low density gas at overdensities of a few .
metals close to galaxies are generally searched for considering the correlation between the metal absorptions observed along a quasar line of sight and the galaxies at matching redshifts present in the field .
all the studies carried out up to now both at high ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and low redshift ( e.g. prochaska et al .
2011 ; tumlinson et al . 2011
; werk et al 2013 ; bordoloi et al . 2014
; liang & chen 2014 ) agree on the significant presence of metals in high and low ionization state at impact parameters at least as large as @xmath21 kpc .
however , this piece of evidence is not sufficient to state that the observed metals were freshly ejected by the associated galaxies , since the same correlation at @xmath22 could originate from the accumulation of older metals in the knots of the cosmic web due to gravitational collapse ( e.g. * ? ? ?
similar results for the size of the regions enriched by galaxies were found based on the clustering properties of metal absorbers ( mainly lines ) along ( e.g. * ? ? ? * ) and across ( e.g. * ? ? ?
* ) the lines of sight to distant quasars .
a few years after the discovery of the first metals polluting the igm , the exploration of the low density regime was attempted using the stacking method @xcite and the individual pixel optical depths ( pod ) of and @xcite .
however , rigorous testing of the analysis procedures by @xcite revealed that in particular the stacking method suffered from severe limitations , leaving the question of whether or not the low density regions of the igm were enriched unanswered . the next step to reach the metallicity level due to a possible pre - enrichment was the observation of an exceptionally high snr spectrum of the gravitationally lensed quasar b1422 + 231 ( @xmath23 ) with the high resolution spectrograph hires at the keck telescope .
* essp00 hereafter ) showed that the @xmath24 detection limit for outside the forest in this object is @xmath25 for a doppler value @xmath26 and the column density distribution function corrected for completeness is fitted by the same power law at least down to @xmath27 , without showing a downturn .
the pod result is consistent with an almost constant @xmath28/@xmath29 down to @xmath30@xmath31 which implies a contribution of absorbers with column densities below the detection limit .
further studies based on the pod technique have been carried out with larger samples @xcite . in particular , (
* s03 hereafter ) based their study on the high snr spectrum of b1422 + 231 plus other 18 objects observed at lower snr .
they used hydrodynamical simulations to convert the measurement of the relation between the optical depths @xmath32 and @xmath33 into a relation between [ c / h ] , overdensity . ] and redshift .
they found that for a fixed overdensity and redshift , the metallicity distribution is close to lognormal with a large scatter ( e.g. @xmath34c / h@xmath35 ) \simeq 0.7 $ ] at @xmath36 and @xmath37 ) ; the median metallicity does not vary significantly with redshift in the range @xmath38 , while it decreases with the overdensity : @xmath39 \simeq -3.47 + 0.65 ( \log \delta -0.5 ) + 0.08(z-3)\ ] ] and the relation is valid down to approximately the mean density .
this result , due to its statistical nature , lacks the information on the distribution and covering factor of metals which could only be derived by a direct detection of the very weak metal lines and their association with the corresponding lines .
after the pioneer work by essp00 , the most ambitious surveys for weak associated metal absorption features ( based on voigt profile fitting of lines ) have detected with @xmath40 associated with absorption with @xmath41 ( e.g. * ? ? ?
* ; * ? ? ?
* ) , corresponding approximately to overdensities of a few at @xmath42 , probing only the densest few per cent of the cosmic volume . at lower column densities , the detection of associated absorption peters out @xcite .
the species is expected to be a more sensitive tracer of metals in the low density igm at @xmath22 ( e.g. * ? ? ?
* ) , but the relevant absorption lines occur in the same wavelength range as the lyman forest , which makes the detection of weak absorption challenging ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
moved by the motivation of directly investigating the metal enrichment of the igm down to the mean density , we have repeated the enterprise of obtaining an ultra - high snr quasar spectrum ( dubbed the _ deep _ spectrum ) using uves at the eso very large telescope , 15 years after the observation of the first and only extremely high snr quasar spectrum of b1422 + 231 . in this paper
, we describe the observation campaign , the data reduction and analysis of the uves deep spectrum of the quasar he0940 - 1050 ( @xmath43 ) in 2 .
section 3 is dedicated to the statistics of the lines and the comparison with previous samples . in 4 , we present the detection rate and discuss the connection with galaxies .
section 5 addresses the core topic , the metallicity of the igm . the final discussion and the conclusions are drawn in 6 .
the computation of the pod for and will be reported in a further paper .
the thermal properties of the igm at @xmath44 have been studied with the forest of the deep spectrum in @xcite . throughout this paper ,
we assume @xmath45 , @xmath46 and @xmath47 .
we carefully chose the target for the deep spectrum to address two main requirements : i ) an emission redshift @xmath48 , representing a trade - off between the need to probe the low - density medium ( at higher redshift the same column density corresponds to lower density ) and that of avoiding the excessive crowding in the forest ( which increases with redshift ) ; ii ) a line of sight free from damped systems , whose presence would significantly decrease the portion of forest available for our study .
the selection was limited to the quasars observed for the uves large programme ( * ? ? ?
* from now on uves lp ) because they were all picked out to be bright , damped - free targets and they have available good quality spectra .
we chose the quasar he0940 - 1050 ( @xmath49 ) which is the brightest quasar of the sample fulfilling our requirements .
aiming at detecting at @xmath1 confidence level lines with column densities @xmath50 ( @xmath51 ) and @xmath52 . ] , we computed the required minimum snr to be @xmath53 and 90 per resolution element in the and forests respectively , using the formula @xcite : @xmath54 where , @xmath55 and @xmath56 are the rest frame wavelength and equivalent width of transition x , respectively , @xmath57 is the speed of light , @xmath58 is the doppler parameter of the lines ( assumed to be @xmath59 ) , and @xmath60 is the velocity resolution element in .
the equivalent width of the line is then converted into a column density assuming the linear regime of the curve of growth : @xmath61 where , @xmath62 is the oscillator strength of transition x. [ tab_obs ] [ cols="<,<,^,^,^,^ , > " , ] to this aim , we have generated a set of simple photonionization models using version 13.0 of , last described by @xcite . in the models , we have assumed a plane parallel slab of gas at a constant temperature @xmath63 , ionized by the inbuilt haardt & madau uv background @xcite at redshift @xmath64 .
the computation stops when a given column density of is reached and the column densities of the studied ions are given in output ( solar relative abundances are assumed ) .
we built a grid of models varying overdensity and metallicity .
the corresponding temperature @xmath63 is defined by the temperature - density relation valid for the igm @xcite : @xmath65 where we assume @xmath66 k and @xmath67 ( following * ? ? ?
the total density is computed based on the mean density at the considered redshift and the considered overdensity .
the column density used to stop the computation is determined using the formula by @xcite : @xmath68 where we have adopted @xmath69 , @xmath70 and @xmath67 , as above .
all the model parameters are reported in table [ cloudy ] . in fig .
[ fig : civ_ovi ] , we plot our observational results for and versus column densities for each absorber , together with the predictions of the photoionization models described above . in the plots ,
we empirically identify three possible regimes . at @xmath4 ( corresponding to @xmath71 at @xmath64 )
, we are probing regions permeated by metals with @xmath72 , likely tracing the cgm close to galaxies .
the range @xmath73 is where observationally we are most sensitive to the very low metallicity gas : what we observe is that @xmath74 per cent of absorbers are enriched to metallicities @xmath75 ( considering upper limits as measurements ) ; while @xmath76 per cent of absorbers are characterized by metallicities @xmath77 , mostly traced by upper limits .
assuming that all upper limits would trace metallicities @xmath78 , a lower limit of 38 per cent of absorbers would have @xmath75 .
the possible detections ( filled pink dots ) and the single detection ( filled blue square ) observed in this column density regime trace a slightly lower metallicity interval ( @xmath79 ) . for @xmath6
our measurements of lines are no longer sensitive to metallicities @xmath80 ( and going down in @xmath81 , the limit rises above @xmath82 ) . as a consequence ,
even if the gas traced by those absorbers was enriched at the same level as the gas traced by larger column densities , our observations could not detect the associated lines .
on the other hand , more stringent upper limits on the metallicity in this overdensity regime can be put using ( lower panel of fig .
[ fig : civ_ovi ] ) for which we find that @xmath83 per cent of the absorbers have metallicities @xmath84 and , in particular , @xmath85 per cent of absorbers have @xmath86 . falling in different metallicity bins , when compared with predictions at @xmath64 .
the hatched histograms correspond to the contribution of upper limits in each metallicity bin .
the grey band is the prediction for the median [ c / h ] abundance computed with eq.uation ( 1 ) by s03 for the considered overdensity range .
see text for more details.,width=340 ] with the caveat that we are dealing with a small sample , we carried out a more detailed analysis of the metallicity of the gas probed by the absorption lines with @xmath73 , range in which our observations are more sensitive to the very low metallicities .
this column density interval , when translated into overdensities with equation ( [ eq : schaye ] ) , corresponds to @xmath87 at @xmath64 . in the considered column density interval ,
42 lines were detected , of which 18 have an associated absorber , while for the remaining 24 lines we have determined an upper limit to @xmath88 . the sample of - pairs has been compared with the results of our models at redshift @xmath64 , computed for the finer metallicity grid : @xmath89 $ ] .
the results of this comparison are shown in form of histogram of the fraction of lines falling in each metallicity bin in fig .
[ fig : metal_hist ] .
the figure shows that the majority of the lines falls in the bin at @xmath90 $ ] , but the majority of the detections falls in the range of metallicities @xmath91 $ ] .
the median of the distribution of lines falls in the bin @xmath92 $ ] .
essp00 estimated the relation between @xmath33 and @xmath93 with the pod method using the extremely high snr spectrum of b1422 + 231 .
using a simple simulation , they showed that the pod result could be reproduced only assuming that more metals than those directly detected were present .
in particular , they assumed that all lines with @xmath17 had an associated line with at least @xmath94 and that all weak lines ( @xmath95 ) were enriched with a constant @xmath96/@xmath97 .
we have verified the validity of those conditions for our sample , starting from the hypothesis that the pod method is a cumulative technique .
thus , we can compute the total column density in corresponding to a certain column density range and divide it by the number of lines to obtain the average column density to be compared with essp00 .
first , all the lines in our sample with @xmath4 have an associated line with @xmath98 . in the range of column densities
@xmath99 , we have computed the average column density to be @xmath100 .
the lower limit is obtained assuming that the lines without a detection have zero and the upper limit considering the upper limits as measurements . on the other hand , for the lines with @xmath101 and adopting the same assumptions described above the average ratio varies between @xmath102/@xmath103 .
this implies that with our data we have directly detected the metals that essp00 found statistically applying the pod method . in fig .
[ fig : metal_hist ] , our result obtained through the direct detection ( or non - detection ) of lines has been compared with the median [ c / h ] abundance derived from the pod analysis by s03 and computed with equation ( [ eq : pod ] ) at the redshift and overdensities reported in the figure .
the two median abundances are in good agreement , if we take into account all the uncertainties derived by the rough modelling adopted in our work and also introduced by the conversion of the observed quantities into the [ c / h ] abundance in s03 . the pod analysis for the spectrum of he0940 - 1050 will be presented in a forthcoming paper ( dodorico et al .
, in preparation ) .
comoving mpc at @xmath64 ( solid line , * ? ? ?
the overplotted curve ( dashed line ) is the result of a fit obtained from the higher resolution runs of @xcite .
see text for more details.per centages are given in bins of @xmath104 = 0.01$].,width=359 ] in order to have a more physical insight on the volume filling factors of the metal enriched gas , we analysed the output of a hydrodynamic simulation run with the gadget - iii code @xcite , presented among the reference set of simulations in @xcite .
this simulation has reasonably well converged flux statistics for the forest and evolve a standard concordance cosmological model with @xmath105 gas and dark matter particles in a periodic box of linear size @xmath106 comoving mpc , while the igm thermal history is in agreement with observational data .
the gas particle distribution extracted at @xmath64 is then interpolated on a grid using a counts in cells scheme that allows us to obtain an overdensity value over a grid of @xmath107 points .
the overdensity values are reported in fig .
[ fig : delta ] in the form of per centage of volume occupied in bins of @xmath108 = 0.01 $ ] ( solid line ) .
however , since the reference simulation that we used is likely to be not perfectly converged in terms of density probability distribution function , we decided to overplot also the results obtained from a higher resolution study by ( * ? ? ? *
dashed curve ) . for the densities of interest here , there is a small but not negligible difference between the two curves and quoted values conservatively bracket the uncertainties between the two results .
[ tab : volume ] r c c c @xmath109 & @xmath110 $ ] & @xmath111 $ ] & @xmath112 $ ] + @xmath113 & @xmath114 $ ] & @xmath115 $ ] & @xmath116 $ ] + + tot .
@xmath117 & @xmath118 & @xmath119 + @xmath120 & @xmath121 & @xmath122 & @xmath123 + @xmath124 & @xmath125 & @xmath126 & @xmath127 + + tot .
& @xmath128 & @xmath129 & @xmath130 + @xmath120 & @xmath131 & @xmath132 & @xmath133 + @xmath124 & @xmath134 & @xmath135 & @xmath136 + in the reference simulation box , the overdensities @xmath137 ( corresponding to the column density range @xmath138 at @xmath64 ) occupy @xmath139 per cent of the volume . using the per centage of absorbers observed at different metallicities
, we estimate that @xmath140 per cent of the volume is enriched at @xmath141 and @xmath142 per cent of the volume has a metallicity @xmath143 .
if we extend our analysis to the mean density ( corresponding in fig .
[ fig : civ_ovi ] to @xmath144 ) then , the probed volume would increase to @xmath145 per cent .
assuming the same per centages of absorbers at the different metallicites hold to this lower densities , the volume enriched above a metallicity @xmath146 becomes @xmath147 per cent .
a conservative upper limit to the enriched volume is determined from our limits on : a maximum of 76 per cent of the gas with overdensities between @xmath148 and @xmath149 ( corresponding to the column density range @xmath150 at @xmath64 ) is enriched to @xmath141 .
this corresponds to an enriched volume fraction of @xmath151 per cent which gives an upper limit of @xmath152 per cent adding the contribution of the larger overdensities .
the higher resolution simulation predicts a smaller volume occupation for the range of overdensities considered in this work . in particular , for the column density range @xmath153 the occupied volume at @xmath64 is 3.0 per cent , for the range @xmath154 is @xmath128 per cent .
these numbers translate into a maximum volume enriched to @xmath141 of @xmath155 per cent .
all the volume filling factors are detailed in table [ tab : volume ] .
it is relevant to note that the contribution of the regions traced by column densities @xmath156 is negligible in terms of enriched volume as can be seen in fig .
[ fig : delta ] .
the association of lines with metal lines is a fundamental step which allows us to estimate the metallicity of the gas at different overdensities which , however , is always affected by a certain degree of uncertainty .
if , for example , the metals are transported in the igm by small metal - rich clumps that then lose their identity in , as suggested by @xcite , many of the metallicities we are deriving could be considered only as average values and the real picture would be much more complicated . with the aim of verifying if the results plotted in figs .
[ fig : civ_ovi ] and [ fig : metal_hist ] are reliable , we have considered only those absorbers for which we had either a detection or an upper limit for both and , disregarding the information on . then , we have also investigated the possible effects of redshift , dividing the obtained sample in two redshift bins : @xmath157 and @xmath158 .
the results are shown in fig .
[ fig : civ_ovi_z ] , where the two redshift subsamples are compared with the predictions of the models described before , computed at the average redshifts of the sub - samples , @xmath159 and @xmath160 , respectively .
the black , dashed lines mark the regions between @xmath161 and 14.8 ( see fig .
[ fig : civ_ovi ] ) , and the column densities have been translated into overdensity values as already described in the previous section .
there is no evidence of a significant difference in the metallicity of the investigated gas in the two considered redshift bins .
we confirm that in the overdensity regime where our observations are most sensitive , from @xmath162 to @xmath163 , most of our measures indicate a metallicity @xmath164 .
in particular , the only two systems for which we have detected both and are at an overdensity of @xmath165 and at a metallicity @xmath166 . considering the total sample ( 88 absorbers ) , we find that @xmath76 per cent of the systems have a metallicity @xmath167 .
this per centage should be considered as a lower limit and would likely increase with the increase of the statistics and of the snr of the spectra .
in this work , we presented the analysis of the ultra - high snr , high resolution spectrum ( dubbed the _ deep _ spectrum ) of the quasar he0940 - 1050 ( @xmath168 ) obtained with the uves spectrograph at the eso vlt telescope .
the aim of this observation was to shed light on the abundance and distribution of metals in the low density intergalactic gas at @xmath169 in order to investigate the details of the enrichment process .
the investigated redshift range , @xmath170 , was chosen in order to have at least the transitions and of the lyman series to allow a reliable determination of @xmath81 . in this whole redshift interval
the deep spectrum is sensitive at @xmath1 to @xmath171 lines with column densities down to @xmath2 ( assuming a doppler value @xmath172 ) , in 90 per cent of this interval the spectrum is sensitive to lines down to @xmath173 and in 60 per cent down to @xmath3 .
we searched also for in same redshift interval .
the search for weak lines at @xmath169 is hindered by the crowded and forests .
indeed , we could detect ( i.e. both lines of doublet were observed ) only three weak absorbers , of which two were identified independently of their association with a absorber .
the observational data were then compared with the results of a grid of models obtained assuming solar chemical composition and varying the metallicity , the overdensity and the temperature of the slab of gas ( see section 5 ) .
the main conclusions that can be drawn from our study are the following .
all lines in our sample with column density @xmath174 show an associated absorption . in the range @xmath175
, 43 per cent of lines has an associated absorption with column density down to our sensitivity limit . at column densities @xmath6
, the detection rates drop to less than 10 per cent .
the comparison with the outcome of our grid of models shows that even if the gas traced by these column densities was as enriched as the one at larger column densities , our spectrum would not have the sensitivity to probe it .
the fraction of lines with @xmath8 showing associated detections in our sample is at least a factor of 2 larger than the fraction of lines in the same column density range , tracing the cgm of relatively bright lyman break galaxies hosted by dark matter haloes of average mass @xmath176 m@xmath10 , as found by ( * ? ? ?
* see fig .
[ fig : civ_detrate ] ) .
this result suggests that metals could lie also at larger distances than 300 physical kpc from the considered galaxies but also that they could be present around lower mass galaxies , in agreement with a variety of theoretical studies ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
\3 . in the range of overdensities in which our data are most sensitive to low metallicities , from @xmath177 to @xmath178 ( corresponding to the range of column densities @xmath179 at @xmath64 ) , the comparison with our grid of models indicates that between @xmath180 and 60 per cent of the absorbers are enriched to metallicities @xmath181 , while the remaining @xmath182 per cent have a metallicity lower than @xmath183 ( see figs .
[ fig : civ_ovi ] and [ fig : metal_hist ] ) . \4 . through a comparison with the predictions of two cosmological simulations
, we derived that the volume filling factor of the igm gas ( down to the mean density ) enriched to a metallicity @xmath11 should be of the order of @xmath12 per cent .
this is again in agreement with predictions by theoretical studies in which the igm was enriched at high redshift by low mass objects . as a final statement
, we would like to underline that the main weakness of this study is that it is based on a single high - resolution quasar spectrum albeit with an exceedingly high signal - to - noise ratio and 500 in the forest at @xmath170 , being larger than @xmath184 in more than 80 per cent of the redshift interval . ] .
such amazing data are the result of a gigantic observational effort performed with the uves spectrograph at the vlt , which will unlikely be repeatable on a short time - scale . as a consequence
it will be difficult to have soon a significative sample of exceptional spectra as this one . in the near future
, the espresso spectrograph @xcite , a second generation vlt instrument whose first light is expected in 2017 , will provide an opportunity for further high snr studies of quasars .
it will offer the unique possibility of working at a resolution @xmath185 collecting the light from all four unit telescopes at the same time , mimicking the outcome of a 16m - equivalent telescope .
for the real breakthrough in this research field , however , we will have to wait 10 - 15 years until the next generation of high - resolution spectrographs working on 30m - class telescopes will be on stage .
based on observations collected at the european southern observatory very large telescope , cerro paranal , chile
programs 166.a-0106 , 079.b-0469 , 185.a-0745 , 092.a-0170 .
we are indebted to an anonymous referee for helpful comments that clarified several points in the paper .
vd is grateful to gwen rudie for sharing and adapting her results on the kbss for fig .
[ fig : civ_detrate ] .
mv is supported by the erc starting grant `` cosmoigm '' and pd51 indark grant .
tsk acknowledges funding support from the erc starting grant `` cosmoigm '' , through grant ga-257670 .
mh was supported by the erc advanced grant 320596 `` the emergence of structure during the epoch of reionisation '' .
pb is supported by the inaf prin-2014 grant `` windy black holes combing galaxy evolution '' . parts of this research were conducted by the australian research council centre of excellence for all - sky astrophysics ( caastro ) , through project number ce110001020 . 99 adelberger k. l. , shapley a. e. , steidel c. c. , pettini m. , erb d. k. , reddy n. a. , 2005 , apj , 629 , 636 aguirre a. , hernquist l. , schaye j. , weinberg d. h. , katz n. , gardner j. 2001 apj , 560 , 599 aracil b. , petitjean p. , pichon c. , bergeron j. , 2004 , a&a , 419 , 811 ballester p. , modigliani a. , boitquin o. , cristiani s. , hanuschik r. , kaufer a. , wolf s. , 2000 , the messenger , 101 , 31 barai p. , viel m. , borgani s. , tescari e. , tornatore l. , dolag k. , killedar m. , monaco p. , dodorico v. , cristiani s. , 2013 , mnras 430 , 3213 becker g. d. , bolton j. s. , haehnelt m. g. , sargent w. l. w. 2011 , mnras 410 , 1096 bergeron j. , et al . , 2004 , the messenger , 118 , 40 bergeron j. , aracil b. , petitjean p. , pichon c. , 2002 , a&a .
396 , l11 bertone s. , stoehr f. , white s. d. m. , 2005 , mnras 359 , 1201 boksenberg a. , sargent w. l. w , 2015 , apjs , 218 , 7 bolton j. s. , becker g.d . , 2009 , mnras , 398 , 1 booth et al .
, 2012 , mnras , 420 , 1053 bordoloi r. , et al . , 2014 apj , 796 , 136 calura f. , matteucci f. , 2006 , mnras , 369 , 465 carswell b. , schaye j. , kim t - s .
, 2002 , apj , 578 , 43 carswell r. f. , webb j. k. , 2014 , astrophysics source code library , record ascl:1408.015 carswell r.f . , webb j.k .
, cooke a. j. , irwin m. j. , 2014 , astrophysics source code library , record ascl:1408.017 cen r. , chisari n. e. , 2011 , apj , 731 , 11 cen r. , miralda - escud j. , ostriker j. p. , rauch m. 1994 , apj , 437 , l9 cicone c. , et al . , 2014 ,
a&a , 562 , 21 cole s. , lacey c. g. , baugh c. m. , frenk c. s. , 2000 , mnras , 319 , 168 cowie l. l. & songaila a. , 1998 , nature 394 , 44 cowie l.l .
, songaila a. , kim t .- s . , hu e.m .
, 1995 , aj , 109 , 1522 cupani g. , et al . , 2015 , mem .
astron . ital . ,
86 , 502 cupani g. , et al . , 2015 ,
adass xxiv , 495 , 289 dekker h. , dodorico s. , kaufer a. , delabre b. , kotzlowski h. , 2000 , proc .
spie , 4008 , 534 dodorico v. , calura f. , cristiani s. , viel m. , 2010 , mnras , 401 , 2715 dodorico v. , bruscoli m. , saitta f. , fontanot f. , viel m. , cristiani s. , monaco p. , 2008 ,
mnras , 389 , 1727 ellison s. l. , songaila a. , schaye j. , pettini m. , 2000 , aj , 120 , 1175 ellison s. l. , lewis g.f . , pettini m. , chaffee f.h . , irwin m.j . , 1999 ,
apj , 520 , 456 ferland g. j. , porter r. l. , van hoof p. a. m. , williams r. j. r. , abel n. p. , lykins m. l. , shaw g. , henney w. j. , stancil p. c. , 2013 , rev .
, 49 , 137 feruglio c. , maiolino r. , piconcelli e. , menci n. , aussel h. , lamastra a. , fiore f. , 2010 , a&a , 518l , 155 fontana a. , ballester p. , 1995
, the messenger , 80 , 37 gonzaga s. , hack w. , fruchter a. , mack j. , eds .
2012 , the drizzlepac handbook .
( baltimore , stsci ) greene j. e. , zakamska n. l. , smith p. s. , 2012 , apj , 746 , 86 haardt f. , madau p. , 2001 in eds . neumann d. m. , van j.t.t . , eds .
, clusters of galaxies and the high redshift universe observed in x - rays : recent results of xmm - newton and chandra , xxxvi rencontres de moriond harrison c. m. , alexander d. m. , mullaney j. r. , swinbank a. m. , 2014 , mnras , 441 , 3306 herbert - fort s. prochaska j. x. , dessauges - zavadsky m. , ellison s. l .. howk j. c. , wolfe a. m. , prochter g. e. , 2006 , pasp , 118 , 1077 hui l. , gnedin n. y. , 1997 , mnras , 292 , 27 katz n. , weinberg d. h. , hernquist l. , 1996 , apjs 105 , 19 keres , d. , katz n. , dav r. , fardal m. , weinberg d. h. , 2009 , mnras , 396 , 2332 liang c. j. , chen , h - w .
, 2014 , mnras , 445 , 2061 lu l. ; sargent w. l. w. , barlow t. a. , rauch m. , 1998 , preprint ( astro - ph/9802189 ) madau p. , ferrara a. , rees m. j. , 2001 , apj , 555 , 92 maiolino r. , et al . , 2012 ,
mnras , 425l , 66 martin c. l. , scannapieco e. , ellison s. l. , hennawi j. f. , djorgovski s. g. , fournier a. p. , 2010 ,
apj , 721 , 174 martin c. l. , 2005 , apj , 621 , 227 misawa t. , charlton j. c. , eracleous m. , 2014 , apj , 792 , 77 molaro p. , et al .
2013 , a&a , 555 , 68 oppenheimer b.d . , dav r. , finlator k. , 2009 , mnras , 396 , 729 pepe f. , et al . , 2014 ,
nachr . , 335 , 8 petitjean p. , mueket j. p. , kates r. e. , 1995 , a&a , 295 , l9 pettini m. , shapley a. e. , steidel c. c. , cuby , j - g . , dickinson m. , moorwood a. f. m. , adelberger k. l. , giavalisco m. , 2001 , apj , 554 , 981 pieri m. m. , schaye j. , aguirre a. , 2006 , apj 638 , 45 porciani c. , madau p. , 2005 , apj 625 , l43 prochaska j. x. , weiner b. , chen h .- w . , mulchaey j. , cooksey k. , 2011 , apj , 740 , 91 rauch m. , 1998 , ara&a , 36 , 267 rorai a. , et al . , 2016 , mnras , submitted rudie g. c. , et al . 2012 apj , 750 , 67 rupke d. s. , veilleux s. , 2011 , apj , 729l , 27 rupke d. s. , veilleux s. , sanders d. b. , 2005 , apjs , 160 , 115 samui s. , subramaniana k. , srianand r. , 2008 , mnras , 385 , 783 scannapieco e. , ferrara a. , madau p. , 2002 ,
apj , 574 , 590 scannapieco e. , pichon c. , aracil b. , petitjean p. , thacker r.j .
, pogosyan d. , bergeron j. , couchman h.m.p . , 2006 , mnras , 365 , 615 schaye j. , carswell r. f. , kim t .- s .
, 2007 , mnras 379 , 1169 schaye j. , aguirre a. , kim t - s . , theuns t. , rauch m. , sargent w.l.w . , 2003 , apj , 596 , 768 ( s03 ) schaye j. 2001 , apj 559 , 507 shapley a. e. , steidel c. c. , pettini m. , adelberger k. l. , 2003 , apj , 588 , 65 shen s. , madau p. , guedes j. , mayer l , prochaska j. x. , wadsley j. 2013 , apj 765 , 89 simcoe r. a. , sargent w. l. w. , rauch m. , 2004 , apj , 606 , 92 somerville r. s. , primack j. r. , 1999 , mnras , 310 , 1087 songaila a. , cowie l.l . , 1996 , aj , 112 , 335 springel v. , 2005 , mnras , 364 , 5 springel v. , hernquist l. , 2003 mnras , 339 , 289 steidel c. c. , erb d. k. , shapley a. e. , pettini m. , reddy n. , bogosavljevi m. , rudie g. c. , rakic o. , 2010 , apj , 717 , 289 sturm e. et al . , 2011 ,
apj , 733l , 16 tescari e. , viel m. , dodorico v. , cristiani s. , calura f. , borgani s. , tornatore l. , 2011 , mnras , 411 , 826 tremonti c. a. , moustakas j. , diamond - stanic a. m. , 2007 , apj , 663l , 77 tumlinson j. et al . , 2011 , science , 334 , 948 turner m. l. , schaye j. , steidel c. c. , rudie g. c. , strom a. l. , 2014 , mnras , 445 , 794 tytler d. , fan x .-
m . , burles s. , cottrell l. , davis c. , kirkman d. , zuo l. , 1995 , in meylan g. , ed . , qso absorption lines , springer - verlag , new york , p. 289
tytler d. , 1987 , apj , 321 , 49 vanzella e. , grazian a. , hayes m. , pentericci l. , schaerer d. , dickinson m. , cristiani s. , giavalisco m. , verhamme a. , nonino m. , rosati p. , 2010
, a&a , 513 , 20 veilleux s. , cecil g. , bland - hawthorn j. , 2005 , ara&a , 43 , 769 viel m. , becker g. d. , bolton j. s. , haehnelt m. g. , 2013 , physrevd , 88 , 043502 weinberg d. et al . , 1999 , in banday a. j. , sheth r. k. , da costa l. n. , eds .
mpa - eso cosmology conference , evolution of large scale structure : from recombination to garching , garching , germany , p.346 weiner b. j. et al .
, 2009 , apj , 692 , 187 werk j. k. , prochaska j. x. , thom c. , tumlinson j. , tripp t. m. , omeara j. m. , peeples m. s. , 2013 , apj , 204 , 17
at @xmath186 corresponding to @xmath187 in the plot
. is not detected at this redshift . ] | in this work , we investigate the abundance and distribution of metals in the intergalactic medium ( igm ) at @xmath0 through the analysis of an ultra - high signal - to - noise ratio uves spectrum of the quasar he0940 - 1050 . in the forest ,
our deep spectrum is sensitive at @xmath1 to lines with column density down to @xmath2 and in 60 per cent of the considered redshift range down to @xmath3 . in our sample ,
all lines with @xmath4 show an associated absorption . in the range @xmath5
, 43 per cent of lines has an associated absorption . at @xmath6
, the detection rates drop to @xmath7 per cent , possibly due to our sensitivity limits and not to an actual variation of the gas abundance properties . in the range @xmath8
, we observe a fraction of lines with detected a factor of 2 larger than the fraction of lines lying in the circum - galactic medium ( cgm ) of relatively bright lyman - break galaxies hosted by dark matter haloes with @xmath9 m@xmath10 .
the comparison of our results with the output of a grid of photoionization models and of two cosmological simulations implies that the volume filling factor of the igm gas enriched to a metallicity @xmath11 should be of the order of @xmath12 per cent . in conclusion ,
our results favour a scenario in which metals are found also outside the cgm of bright star - forming galaxies , possibly due to pollution by lower mass objects and/or to an early enrichment by the first sources .
[ firstpage ] galaxies : abundances - intergalactic medium - quasars : absorption lines - cosmology : observations . |
the isometry problem or more specifically the isometrical deformation problem of surfaces in a 3-dimensional euclid space form preserving mean curvature has been studied by a number of mathematicians .
one of the first mathematicians who has studied this subject is bonnet @xcite .
bonnet , after pointing out that the isometry problem of a surface preserving mean curvature can not be solved in the general case , showed that all surfaces with constant mean curvature can be isometrically mapped to each other and deformable surfaces with non constant mean curvature are isothermic weingarten surfaces which can be deformable to the revolution surfaces .
cartan @xcite considered the problem as an application of differential forms and after rather long calculations , classified weingarten surfaces which can be deformable to the revolution surfaces and showed that there are finite number surfaces in each class .
recently , this problem has been reconsidered .
chern @xcite , using differential forms , obtained a characterization for an isometrical deformation preserving mean curvature .
roussos @xcite who has many works on this subject , using chern s method , in the general case , obtained a characterization for the isometry preserving mean curvature .
some of the important contributions to this subject are by voss @xcite , bobenko and either @xcite , kenmotsu @xcite , colares and kenmotsu @xcite , roussos @xcite , xiuxiong and chia - kuei@xcite .
the surface which admits an isometry - preserving mean curvature is called bonnet surface .
one of the bonnet surfaces which can be isometrically mapped to each other is called bonnet associate of the other surface .
similar definitions are valid for the bonnet hypersurfaces .
kokubu @xcite considered deformable bonnet hypersurfaces and showed that they consist of bonnet surfaces in @xmath3 and bonnet hypersurfaces in @xmath4 .
later , soyuok @xcite examined the problem of determining bonnet surfaces in @xmath3 and showed that a surface is a bonnet surface if and only if a surface has a special isothermal parameter system .
moreover , as an application of this , he obtained cartan s results and explained geometrically why there are a finite number of surfaces in the classes determined by cartan .
soyuok @xcite , in another work , examined the problem determining bonnet hypersurfaces in @xmath4 and obtained quantities of these hypersurfaces .
furthermore , he showed that a 3-dimensional hypersurface is a bonnet hypersurface if and only if it has a special orthogonal net .
let us consider @xmath5 and @xmath6 to be non - minimal bonnet associate hypersurfaces with no umbilical points in @xmath0 .
therefore , the ranks of their shape operators are the same and less than 3 @xcite
. let the local orthonormal frame fields of @xmath5 and @xmath6 consist of the principal vectors @xmath7 and @xmath8 where corresponding principal curvatures are @xmath9 and @xmath10 respectively . since , the null spaces of @xmath5 and @xmath6 are the same at the corresponding points @xcite , @xmath11 & \text{span}\{e_{3},e_{4},\dots , e_{n}\ } = \text{span}\{e_{3}^{\prime},e_{4}^{\prime},\dots , e_{n}^{\prime}\}.\end{aligned}\ ] ] because @xmath5 and @xmath6 are associate to each other , the first and the second of the curvature lines of @xmath6 correspond to the orthogonal curves on @xmath5 .
these curves are called bonnet curves of @xmath5 .
let @xmath12 and @xmath13 be the unit vectors of the bisectors of bonnet curves with curvature lines ( first and second ) of @xmath5 .
we take @xmath14 as the orthonormal frame field for @xmath5 and we can also choose @xmath15 as the orthonormal frame field for @xmath6 .
let us denote the shape operator and the riemannian connection of @xmath5 by @xmath16 and @xmath17 respectively .
then @xmath18 ( k_{1}-k_{2})\sin \theta & k_{1}(1-\cos\theta)+k_{2}(1+\cos \theta ) & 0 & \dots & 0\\[.2pc ] 0 & 0 & 0 & \dots & 0\\[.2pc ] \dots & \dots & \dots & \dots & \dots\\[.2pc ] 0 & 0 & 0 & \dots & 0 \end{array } \right)\nonumber\\\end{aligned}\ ] ] can be written to the frame field @xmath19 where @xmath20 is the angle between @xmath21 and @xmath22 . because the angle between @xmath22 and @xmath12 is @xmath23 , the shape operator of @xmath6 with respect to the same frame field
can be written as @xmath24 -(k_{1}-k_{2})\sin \theta & k_{1}(1\!-\!\cos\theta)+k_{2}(1\!+\!\cos\theta ) & 0 & \dots & 0\\[.2pc ] 0 & 0 & 0 & \dots & 0\\[.2pc ] \dots & \dots & \dots & \dots & \dots\\[.2pc ] 0 & 0 & 0 & \dots & 0 \end{array } \right).\nonumber\\\end{aligned}\ ] ] we shall use the following convention on the ranges of indices in this work : @xmath25 3 & \leq p , q , r,\ldots \leq n,\\[.2pc ] 1 & \leq a , b , c,\ldots \leq n.\end{aligned}\ ] ]
let us define the functions @xmath26 and @xmath27 as follows : @xmath28 \bar{k } & = \{k_{1}(1-\cos \theta)+k_{2}(1+\cos \theta)\}=h - j\cos \theta,\nonumber\\[.4pc ] t & = \frac{1}{2}(k_{1}-k_{2})\sin \theta = j\sin \theta,\end{aligned}\ ] ] where @xmath29 is a mean curvature and @xmath30 the codazzi equation @xcite is given by @xmath31 accordingly , using the shape operator of @xmath5 and the above definitions we can write the codazzi equations of @xmath5 in the form : @xmath32 & e_{1}(\bar{k})+(\bar{k } -k)w_{1}^{2}(e_{2})=e_{2}(t)+2tw_{2}^{1}(e_{1}),\nonumber\\[.2pc ] & \bar{k}w_{2}^{p}(e_{1})-kw_{1}^{p}(e_{2})=t[w_{2}^{p}(e_{2})-w_{1}^{p}(e_{1})],\nonumber\\[.2pc ] & e_{p}(k)+kw_{p}^{1}(e_{1})=t[2w_{1}^{2}(e_{p})-w_{p}^{2}(e_{1})],\nonumber\\[.2pc ] & e_{p}(t)+tw_{p}^{1}(e_{1})=\bar{k}[w_{1}^{2}(e_{p})-w_{p}^{2}(e_{1 } ) ] -kw_{1}^{2}(e_{p}),\nonumber\ ] ] @xmath33 -\bar{k}w_{2}^{1}(e_{p}),\nonumber\\[.2pc ] & e_{p}(\bar{k})+\bar{k}w_{p}^{2}(e_{2})=t[2w_{2}^{1}(e_{p})-w_{p}^{1}(e_{2})],\nonumber\\[.2pc ] & kw_{1}^{p}(e_{p})=-tw_{2}^{p}(e_{p}),\nonumber\\[.2pc ] & kw_{1}^{q}(e_{p})=-tw_{2}^{q}(e_{p}),\nonumber\\[.2pc ] & \bar{k}w_{2}^{p}(e_{p})=-tw_{1}^{p}(e_{p}),\nonumber\\[.2pc ] & k[w_{p}^{1}(e_{q})-w_{q}^{1}(e_{p } ) ] = -t[w_{p}^{2}(e_{q})-w_{q}^{2}(e_{p})],\nonumber\\[.2pc ] & \bar{k}[w_{p}^{2}(e_{q})-w_{q}^{2}(e_{p } ) ] = -t[w_{p}^{1}(e_{q})-w_{q}^{1}(e_{p})],\end{aligned}\ ] ] where @xmath34 are the connection forms .
because of ( 2.1 ) , ( 2.2 ) and ( 3.1 ) , in order to find the codazzi equations of @xmath6 it is enough to replace @xmath27 by @xmath35 in ( 3.4 )
. therefore , from the codazzi equations of @xmath5 and @xmath6 the following equations are obtained : @xmath36 & w_{1}^{q}(e_{p})=w_{2}^{q}(e_{p})=0,\nonumber\\[.2pc ] & w_{p}^{2}(e_{2})=w_{p}^{1}(e_{1}),\nonumber\\[.2pc ] & w_{p}^{1}(e_{2})=-w_{p}^{2}(e_{1})=2w_{2}^{1}(e_{p}),\nonumber\\[.2pc ] & hw_{p}^{1}(e_{2})=0,\nonumber\\[.2pc ] & e_{p}(k)+kt_{p}=0,\nonumber\\[.2pc ] & e_{p}(\bar{k})+\bar{k}t_{p}=0,\nonumber\\[.2pc ] & e_{p}(t)+tt_{p}=0,\nonumber\\[.2pc ] & e_{2}(k)+(k-\bar{k})h=0,\nonumber\\[.2pc ] & e_{1}(\bar{k})+(\bar{k}-k)\bar{h}=0,\nonumber\\[.2pc ] & e_{1}(t)+2t\bar{h}=0,\nonumber\\[.2pc ] & e_{2}(t)+2th=0,\end{aligned}\ ] ] where the functions @xmath37 , @xmath38 and @xmath39 are defined as @xmath40 & h = w_{2}^{1}(e_{1 } ) , \quad \bar{h}=w_{1}^{2}(e_{2}).\end{aligned}\ ] ] since we assume that our hypersurface is a non - minimal one , @xmath41 and from the fifth equation of ( 3.5 ) we get @xmath42 from the fourth equation of ( 3.5 ) , @xmath43 are obtained .
therefore , we have the following lemma
. a bonnet hypersurface must satisfy the codazzi equations , @xmath44 & e_{p}(\bar{k})+\bar{k}t_{p}=0,\nonumber\\[.2pc ] & e_{p}(t)+tt_{p}=0,\nonumber\\[.2pc ] & e_{2}(k)+(k-\bar{k})h=0,\nonumber\\[.2pc ] & e_{1}(\bar{k})+(\bar{k}-k)\bar{h}=0,\nonumber\\[.2pc ] & e_{1}(t)+2t\bar{h}=0,\nonumber\\[.2pc ] & e_{2}(t)+2th=0.\end{aligned}\ ] ] the gauss equations of @xmath5 hypersurface @xcite are given by the equations @xmath45}^{e_{c}}\\[.3pc ] & \quad\ , = \langle l(e_{b}),e_{c}\rangle l(e_{a})-\langle l(e_{a}),e_{c}\rangle l(e_{b}).\end{aligned}\ ] ] using ( 3.7 ) and ( 3.8 ) in the gauss equations of @xmath5 , we have the following lemma . the gauss equations of @xmath5 can be written as follows : @xmath46 & e_{p}(h)+t_{p}h=0,\nonumber\\[.2pc ] & e_{p}(\bar{h})+t_{p}\bar{h}=0,\nonumber\\[.2pc ] & e_{p}(t_{q})-w_{q}^{r}(e_{p})t_{r}+t_{p}t_{q}=0,\nonumber\\[.2pc ] & e_{1}(t_{p})-w_{p}^{q}(e_{1})t_{q}=0,\nonumber\\[.2pc ] & e_{2}(t_{p})-w_{p}^{q}(e_{2})t_{q}=0,\nonumber\\[.2pc ] & e_{p}(w_{r}^{s}(e_{q } ) ) + w_{r}^{t}(e_{q})w_{t}^{s}(e_{p})-e_{q}(w_{r}^{s}(e_{p}))\nonumber\\[.2pc ] & \quad\ , -w_{r}^{t}(e_{p})w_{t}^{s}(e_{q})-w_{\!p}^{t}(e_{q } ) w_{r}^{s}(e_{t})+w_{q}^{t}(e_{p})w_{r}^{s}(e_{t})=0.\end{aligned}\ ] ] let @xmath47 be the dual frame field of @xmath48 .
thus , by means of ( 3.6 ) and ( 3.7 ) @xmath49 w_{p}^{1 } & = t_{p}w^{1},\\[.2pc ] w_{p}^{2 } & = t_{p}w^{2},\\[.2pc ] w_{p}^{q } & = w_{p}^{q}(e_{a})w^{a}\end{aligned}\ ] ] are obtained .
accordingly , the structure equations @xmath50 can be written as follows : @xmath51 \hbox{d}w^{2 } & = \bar{h}w^{1}\wedge w^{2}-t_{p}w^{2}\wedge w^{p},\nonumber\\[.2pc ] \hbox{d}w^{q } & = -w_{p}^{q}(e_{a})w^{a}\wedge w^{p}.\end{aligned}\ ] ] now , we consider the differential @xmath52 of @xmath53 with respect to the dual frame field @xmath47 .
using ( 3.7 ) and ( 3.10 ) in @xmath54 , we obtain the compatibility equations @xmath55 & f_{1p}+t_{p}f_{1}-f_{p1}+f_{q}w_{p}^{q}(e_{1})=0,\nonumber\\[.2pc ] & f_{2p}+t_{p}f_{2}-f_{p2}+f_{q}w_{p}^{q}(e_{2})=0,\nonumber\\[.2pc ] & f_{rp}+t_{p}f_{r}-f_{pr}+f_{q}w_{p}^{q}(e_{r})=0.\end{aligned}\ ] ] the sixth and seventh equations of the system ( 3.7 ) can be written in the form @xmath56 so , using the first equation of ( 3.11 ) , @xmath57
let @xmath58 be the components of the first fundamental tensor .
the form @xmath59 can be written as @xmath60 in the suitable system of local coordinates @xmath61 @xcite .
so , from the structure equations ( 3.10 ) , we get @xmath62 & w_{r}^{a}(e_{p})=0,\\[.2pc ] & g_{pp}=c^{p}(x^{p}),\\[.2pc ] & h=\frac{(\sqrt{g_{11 } } ) _ { x^{2}}}{\sqrt{g_{11}g_{22}}}=(\ln \sqrt{g_{11}})_{2},\quad \bar{h}= \frac{(\sqrt{g_{22}})_{x^{1}}}{\sqrt{g_{11}g_{22}}}=(\ln \sqrt{g_{22}})_{1},\\[.2pc ] & t_{p}=\frac{(\sqrt{g_{11}})_{x^{p}}}{\sqrt{g_{11}g_{pp}}}= \frac{(\sqrt{g_{22}})_{x^{p}}}{\sqrt{g_{22}g_{pp}}},\end{aligned}\ ] ] where @xmath63 because of ( 4.4 ) in ( 3.12 ) , equation ( 3.13 ) is equivalent to @xmath64 therefore , we have @xmath65 where @xmath66 and @xmath67 are arbitrary functions .
let us introduce the new coordinates by means of the following scaling transformation : @xmath68 then , from ( 4.7 ) and ( 4.3 ) we get @xmath69 here we have again denoted the new coordinates by @xmath70 . accordingly ( 4.5 ) reduces to @xmath71 because of ( 4.1 ) the fifth and sixth equations of the gauss equations ( 3.9 ) take the forms @xmath72 so , @xmath37 does not depend on @xmath73 and @xmath74 .
moreover , using ( 4.2 ) , the fourth equation of ( 3.8 ) can be written as @xmath75 and so @xmath76 now , let us consider the following two cases : 1 .
some @xmath37 s are not zero
all @xmath37 s are zero .
assume that some @xmath37 s are not zero .
let the number of non zero @xmath37 s be @xmath77 .
renewing the indices @xmath78 of non zero @xmath37 s we can choose them as @xmath79 .
hence we can take the non zero @xmath37 s to be @xmath80 @xmath81 . solving the equations ( 4.10 ) and ( 4.11 )
, we find @xmath82 where @xmath83 are non zero constants of integration .
accordingly , from ( 4.9 ) we have @xmath84 where @xmath85 is an arbitrary function . by means of ( 4.4 ) and ( 4.13 ) , the second and the third gauss equations of ( 3.9 ) are automatically satisfied using ( 4.8 ) and ( 4.9 ) , let us solve the third equation of ( 3.7 ) and then let us solve eqs .
( 3.12 ) , ( 4.4 ) and ( 4.13 ) .
thus @xmath86 where @xmath87 is a constant . on the other hand ,
using ( 4.8 ) and ( 4.12 ) , solutions of the first and the second equations of the system ( 3.7 ) are obtained as @xmath88 now , let us find the mean curvature @xmath29 .
since @xmath89 , according to ( 3.1 ) we have @xmath90 moreover , from ( 3.1 ) , ( 4.15 ) and ( 4.16 ) , we have @xmath91 and @xmath92 all of codazzi s equations ( 3.8 ) are satisfied except the fourth and the fifth equation and all of gauss equations ( 3.8 ) are satisfied except the first equation .
now , we have to consider these three equations . by using ( 3.1 ) ,
( 4.4 ) , ( 4.6 ) , ( 4.14 ) , ( 4.16 ) , ( 4.17 ) and ( 4.18 ) , the fourth and the fifth equations of codazzi s equations ( 3.8 ) can be written as follows : @xmath93 the first equation of gauss s equations ( 3.8 ) is reduced to the following equation : @xmath94 & \quad\ , = 2 ( \mathfrak{h}^{2}-\mathfrak{j}^{2}).\end{aligned}\ ] ]
let us denote the components of the second fundamental tensor with @xmath95 .
they are given by @xcite @xmath96 thus , we get @xmath97 by using ( 4.13 ) , ( 4.14 ) and ( 4.15 ) , non zero components can be written as @xmath98 b_{22 } & = ( c^{3}x^{3}+\dots + c^{r}x^{r } ) \frac{(\mathfrak{h}-\mathfrak{j } \sin \theta)}{\vert \mathfrak{j}\vert \sin \theta},\\[.3pc ] b_{12 } & = c(c^{3}x^{3}+\dots + c^{r}x^{r}).\end{aligned}\ ] ] we can make a suitable scale transformation @xmath99 .
thus , the components of the first and the second fundamental tensors of bonnet hypersurface @xmath5 in an orthogonal coordinate system @xmath100 are obtained as @xmath101 b_{11 } & = ( c^{3}x^{3}+\dots + c^{r}x^{r } ) \frac{(\mathfrak{h}+\mathfrak{j } \sin \theta)}{\vert \mathfrak{j}\vert \sin \theta},\nonumber\\[.3pc ] b_{22 } & = ( c^{3}x^{3}+\dots + c^{r}x^{r } ) \frac{(\mathfrak{h}-\mathfrak{j } \sin \theta)}{\vert \mathfrak{j}\vert \sin \theta},\nonumber\\[.3pc ] b_{12 } & = \in ( c^{3}x^{3}+\dots + c^{r}x^{r}),\quad \in = \pm 1\nonumber\\[.3pc ] b_{iq } & = b_{pq}=0,\quad \mathfrak{h}=\mathfrak{h}(x^{1},x^{2}),\dots , \mathfrak{j}=\mathfrak{j}(x^{1},x^{2}).\end{aligned}\ ] ] because of ( 2.1 ) and ( 2.2 ) , the components of the first and the second fundamental tensors of associate bonnet hypersurface @xmath6 can be written in the form : @xmath102 b_{11}^{\prime } & = ( c^{3}x^{3}+\dots + c^{r}x^{r } ) \frac{(\mathfrak{h}+ \mathfrak{j}\sin \theta)}{\vert \mathfrak{j}\vert \sin \theta},\nonumber\\[.3pc ] b_{22}^{\prime } & = ( c^{3}x^{3}+\dots + c^{r}x^{r})\frac{(\mathfrak{h}-
\mathfrak{j}\sin \theta)}{\vert \mathfrak{j}\vert \sin\theta},\nonumber\\[.3pc ] b_{12}^{\prime } & = -\in ( c^{3}x^{3}+\dots + c^{r}x^{r}),\quad \in = \pm 1\nonumber\\[.3pc ] b_{iq}^{\prime } & = b_{pq}^{\prime}=0.\end{aligned}\ ] ] it is seen that all the fundamental quantities of @xmath5 and @xmath6 are the same , except @xmath103 and @xmath104 .
there is the relation @xmath105 between @xmath103 and @xmath104 .
let us consider the case where all @xmath106 .
then , it is easily seen that in order to obtain the quantities in this case , it is sufficient to replace @xmath107 by @xmath108 and @xmath109 by @xmath110 in the quantities of case @xmath108 .
thus , in this case , the fundamental quantities of @xmath5 and @xmath6 in an orthogonal coordinate system @xmath100 are respectively given by @xmath111 b_{11 } & = \frac{(\mathfrak{h}+\mathfrak{j}\sin \theta)}{\vert \mathfrak{j}\vert \sin \theta},\nonumber\\[.3pc ] b_{22 } & = \frac{(\mathfrak{h}-\mathfrak{j}\sin \theta ) } { \vert \mathfrak{j}\vert \sin \theta},\nonumber\\[.3pc ] b_{12 } & = \in,\quad \in = \pm 1\nonumber\\[.3pc ] b_{iq } & = b_{pq}=0,\quad \mathfrak{h}=\mathfrak{h}(x^{1},x^{2}),\dots,\quad \mathfrak{j}=\mathfrak{j}(x^{1},x^{2})\end{aligned}\ ] ] and @xmath112 b_{11}^{\prime } & = \frac{(\mathfrak{h}+\mathfrak{j}\sin \theta)}{\vert \mathfrak{j}\vert \sin \theta},\nonumber\\[.3pc ] b_{22}^{\prime } & = \frac{(\mathfrak{h}-\mathfrak{j}\sin \theta)}{\vert \mathfrak{j}\vert \sin \theta},\nonumber\\[.3pc ] b_{12}^{\prime } & = -\in,\quad \in = \pm 1\nonumber\\[.3pc ] b_{iq}^{\prime } & = b_{pq}^{\prime}=0.\end{aligned}\ ] ] @xmath113 if the fundamental quantities in an orthogonal coordinate system @xmath100 are in the form @xmath114 or @xmath115 where @xmath116 are constants .
the net which consists of this orthogonal coordinate system is called @xmath2-net .
therefore , we have the following theorem .
a non - minimal hypersurface in @xmath0 with no umbilical points is a bonnet hypersurface if and only if it has an @xmath2-net .
moreover , the fundamental quantities of the bonnet hypersurface and its bonnet associate are given by @xmath117 and @xmath118 respectively or @xmath119 and @xmath120 respectively .
the authors would like to thank markus roussos for his interest in this work .
chern s s , deformations of surfaces preserving principal curvature , differantial geometry and complex analysis , h. e. rauch memorial volume ( eds ) i chavel and h m farkas ( 1985 ) ( springer - verlag ) pp . | the problem of determining the _ bonnet hypersurfaces in _ @xmath0 , for @xmath1 , is studied here .
these hypersurfaces are by definition those that can be isometrically mapped to another hypersurface or to itself ( as locus ) by at least one nontrivial isometry preserving the mean curvature . the other hypersurface and/or ( the locus of ) itself is called _ bonnet associate _ of the initial hypersurface . the orthogonal net which is called
is special and very important for our study and it is described on a hypersurface .
it is proved that , non - minimal hypersurface in @xmath0 with no umbilical points is a bonnet hypersurface if and only if it has an@xmath2-net .
[ theore]*theorem * |
in this work , we study the stability of the jump - diffusion it s stochastic differential equations ( sdes ) of the form @xmath0,\,\ , t>0.\end{aligned}\ ] ] here @xmath1 is a @xmath2-dimensional brownian motion , @xmath3 , @xmath4 satisfies the one - sided lipschitz condition and the polynomial growth condition , the functions @xmath5 and @xmath6 satisfy the globally lipschitz condition , and @xmath7 is a one dimensional poisson process with parameter @xmath8 .
the one - sided lipschitz function @xmath9 can be decomposed as @xmath10 , where the function @xmath11 is the global lipschitz continuous part and @xmath12 is the non - global lipschitz continuous part . using this decomposition , we can rewrite the jump - diffusion sdes in the following equivalent form @xmath13 equations of type arise in a range of scientific , engineering and financial applications ( see @xcite and references therein ) .
the standard explicit methods for approximating sdes of type is the euler - maruyama method and implicit schemes @xcite .
their numerical analysis have been studied in @xcite with implicit and explicit schemes .
recently it has been proved ( see @xcite ) that the euler - maruyama method often fails to converge strongly to the exact solution of nonlinear sdes of the form without jump term when at least one of the functions @xmath9 and @xmath14 grows superlinearly . to overcome this drawback of the euler - maruyama method , numerical approximation which computational cost is close to that of the euler - maruyama method and which converge strongly even in the case the function @xmath9 is superlinearly growing
was first introduced in @xcite . in our accompanied paper @xcite
, the work in @xcite has been extended to sdes of type and the strong convergence of the following numerical schemes has been investigated @xmath15 and @xmath16 where @xmath17 is the time step - size , @xmath18 is the number of time subdivisions , @xmath19 and @xmath20 .
the scheme is called the non compensated tamed scheme ( ncts ) , while scheme is called the semi - tamed scheme .
strong and weak convergences are not the only features of numerical techniques .
stability for sdes is also a good feature as the information about step size for which does a particular numerical method replicate the stability properties of the exact solution is valuable .
the linear stability is an extension of the deterministic a - stability while exponential stability can guarantee that errors introduced in one time step will decay exponentially in future time steps , exponential stability also implies asymptotic stability @xcite . by the chebyshev inequality and the borel
cantelli lemma , it is well known that exponential mean square stability implies almost sure stability @xcite . the stability of classical implicit and explicit methods for
are well understood @xcite .
although the strong convergence of the ncts and sts schemes given respectively by and have been provided in @xcite , a rigorous stability properties have not yet investigated to the best of our knowledge .
the goal of this paper is to study the linear stability and the exponential stability of and for sdes driven by both brownian motion and poisson jump .
our study will also provide the rigorous study of linear stabilities of schemes and for sdes without jump , which have not yet studied to the best of our knowledge .
the paper is organised as follows . the linear mean - square stability and the exponential mean - square stability of the tamed and semi - tamed schemes are investigated respectively in section [ linearsta ] and section [ nlinearsta ] .
section [ simulations ] presents numerical simulations to sustain the theoretical results .
we also compare the stability behaviors of tamed and semi - tamed schemes with those of backward euler and split - step backward euler , this comparison shows the good behavior of the semi - tamed scheme and therefore confirms the previous study in @xcite for sdes without jump .
throughout this work , @xmath21 denotes a complete probability space with a filtration @xmath22 . for all @xmath23 ,
we denote by @xmath24 , @xmath25 and @xmath26 , for all @xmath27 .
the goal of this section is to find the time step - size limit for which the tamed euler scheme and the semi - tamed euler scheme are stable in the linear mean - square sense . for the scalar linear test problem
, the concept of a - stability of a numerical method may be interpreted as problem stable @xmath28 method stable for all @xmath29 .
we consider the following linear test equation with real and scalar coefficients .
@xmath30 where @xmath31 satisfied @xmath32 .
it is proved in @xcite that the exact solution of is mean - square stable if and only if @xmath33 using the discrete form of , the numerical schemes and will be therefore mean - square stable if @xmath34 and @xmath35 the following result provides the time step - size limit for which the semi - tamed scheme ( sts ) is is mean - square stable .
assume that @xmath34 , then @xmath36 and the semi - tamed scheme is mean - square stable if and only if @xmath37 applying the semi - tamed euler scheme to leads to @xmath38 squaring both sides of leads to @xmath39 taking expectation in both sides of and using the relations @xmath40 , @xmath41 and @xmath42 with the fact that @xmath43 and @xmath44 are independents leads to @xmath45 so , the semi - tamed scheme is stable if and only if @xmath46 that is @xmath47 .
the following result provide the time step - size limit for which the non compensated tamed scheme ( ncts ) is stable .
[ thmncts ] assume that @xmath34 , the tamed euler scheme is mean - square stable if one of the following conditions is satisfied @xmath48 @xmath49 , @xmath50 and @xmath51 .
+ @xmath48 @xmath52 and @xmath47 . applying the tamed euler scheme to equation leads to @xmath53 by squaring both sides of leads to @xmath54 using the inequality @xmath55 ,
the previous equality becomes @xmath56 taking expectation in both sides of the previous equality and using independence and the fact that @xmath57 , @xmath40 , @xmath58 , @xmath59 leads to : @xmath60\mathbb{e}|x_n|^2\nonumber\\ & + & \mathbb{e}\left(\dfrac{2ax^2_n\delta t(1+\lambda c\delta t)}{1+\delta
t|ax_n|}\right ) . \label{ch5eq2}\end{aligned}\ ] ] @xmath48 if @xmath49 , it follows from that @xmath61\delta t\}\mathbb{e}|x_n|^2.\end{aligned}\ ] ] therefore , the numerical solution is stable if @xmath62\delta t<1.\end{aligned}\ ] ] that is @xmath51 .
@xmath48 if @xmath52 , using the fact that @xmath63 , inequality becomes @xmath64\mathbb{e}|x_n|^2 .
\label{ch5eq3}\end{aligned}\ ] ] therefore , it follows from that the numerical solution is stable if @xmath65 .
that is @xmath47 . in theorem [ thmncts ]
, we can easily check that if @xmath34 , we have : @xmath66 , c\geq 0,\\ \delta t < \dfrac{2a - l}{a^2+\lambda^2 c^2}\\ \end{array } \right . \bigcup \left\lbrace \begin{array}{l
} a \in ( l/2,0 ) , c<0,\\ \delta t < \dfrac{2a - l}{a^2+\lambda^2 c^2 } \\ \delta t\leq \dfrac{-1 } { \lambda c } \\ \end{array } \right.\\ \bigcup \left\lbrace \begin{array}{l } a>0 , c<0 \\
\delta t < \dfrac{2a - l}{a^2+\lambda^2 c^2 } \\
\delta t \geq \dfrac{-1 } { \lambda c } \end{array } \right . \end{aligned}\ ] ] @xmath67
in this section , we focus on the exponential mean - square stability of the approximation .
we follow closely @xcite and assume that @xmath68 and @xmath69 .
it is proved in @xcite that under the following conditions , @xmath70 for all @xmath71 , where @xmath72 , @xmath73 and @xmath74 are constants , the exact solution of sde is nonlinear mean - square stable if @xmath75 .
indeed under the above assumptions , we have ( * ? ? ?
* theorem 4 ) @xmath76 so , if @xmath75 we have @xmath77 and the exact solution @xmath78 is exponentially mean - square stable . in the sequel of this section , we will use some weak assumptions , which of courses imply that the conditions - hold . in order to study the nonlinear stability of the semi - tamed scheme ( sts ) ,
we make also the following assumptions [ ch5assumption2 ] there exist some positive constants @xmath79 , @xmath80,@xmath81 , @xmath82 , @xmath83 , @xmath84 and @xmath85 such that @xmath86 we denote by @xmath87 and we will always assume that @xmath88 to ensure the stability of the exact solution . the nonlinear stability of sts scheme is given in the following theorem .
[ nt1 ] under assumptions [ ch5assumption2 ] and the further hypothesis @xmath89 , for any stepsize @xmath90\overline{\beta}}\wedge\dfrac{2\beta-\overline{\beta}}{2(k+\lambda c)\overline{\beta}}$ ] , there exists a constant @xmath91 such that @xmath92 and the numerical solution is exponentiallly mean - square stable .
the numerical solution is given by @xmath93 where @xmath94 .
+ taking the inner product in both sides of the previous equation leads to @xmath95 using assumptions [ ch5assumption2 ] , it follows that @xmath96 @xmath97 set @xmath98 .
@xmath48 on @xmath99 we have @xmath100 therefore using , and in yields @xmath101\|y_n\|^{a+1}}{1+\delta t\|v(y_n)\|}. \label{ch5meansemi4a } \end{aligned}\ ] ] for @xmath102 , which is equivalent to @xmath103 , becomes @xmath104 @xmath48 on @xmath105 we have @xmath106 therefore , using , and in yields @xmath107\|y_n\|^{a+1}}{1+\delta t \|v(y_n)\|}. \label{ch5meansemi5a } \end{aligned}\ ] ] for @xmath108\overline{\beta}}$ ] , which is equivalent to @xmath109 , becomes @xmath110 finally , from the discussion above on @xmath99 and @xmath105 , it follows that on @xmath111 , if @xmath112\overline{\beta}}\wedge\dfrac{2\beta-\overline{\beta}}{2(k+\lambda c)\overline{\beta}}$ ] then we have @xmath113 taking the expectation in both sides of and using the martingale properties of @xmath43 and @xmath114 leads to @xmath115 from assumptions [ ch5assumption2 ] , we have @xmath116 so inequality gives @xmath117\mathbb{e}\|y_n\|^2 .
\end{aligned}\ ] ] iterating the previous inequality leads to @xmath118^n\mathbb{e}\| x_0\|^2 \end{aligned}\ ] ] for @xmath112\overline{\beta}}\wedge\dfrac{2\beta-\overline{\beta}}{2(k+\lambda c)\overline{\beta}}$ ] . the stability occurs if and only if @xmath119 , so we should also have @xmath120 that is @xmath121}{(k+\lambda c)^2 } = \dfrac{-\alpha_1}{(k+\lambda c)^2 } , \end{aligned}\ ] ] and there exists a constant @xmath91 such that @xmath122^n\mathbb{e}\| x_0\|^2\leq \mathbb{e}\|x_0\|^2 e^{-\gamma t_n } , \end{aligned}\ ] ] by the taylor expansion , as @xmath123 we obviously have @xmath124 . in order to analyse the nonlinear mean - square stability of the tamed euler scheme ( ncts ) , we use the following assumption . [ ch5assumption3 ] there exist some positive constant @xmath80 , @xmath81 , @xmath84 , @xmath72 , @xmath82 , @xmath79 , @xmath83 and @xmath85 such that : @xmath125 apart from , assumption [ ch5assumption3 ] is a consequence of assumption [ ch5assumption2 ] . using assumption [ ch5assumption3 ]
, we can easily check that the exact solution of is exponentiallly mean - square stable if @xmath126 . under assumption [ ch5assumption3 ] ,
if @xmath127 , @xmath128 , and @xmath129 , for any stepsize @xmath130 there exists a constant @xmath91 such that @xmath131 and the numerical solution is exponentiallly mean - square stable . from equation , we have @xmath132 using assumption [ ch5assumption3 ] , it follows that @xmath133 so from assumption [ ch5assumption3 ] , we have @xmath134 let us define @xmath135 .
@xmath48 on @xmath99 , using assumption [ ch5assumption3 ] we have : @xmath136 therefore using and in yields @xmath137\|x_n\|^{a+1}}{1+\delta t\|f(x_n)\|}. \label{ch5meantamed6 } \end{aligned}\ ] ] since @xmath128 , becomes @xmath138 @xmath48 on @xmath105 , using assumption [ ch5assumption3 ] and the inequality @xmath139 we have @xmath140 therefore , using and , becomes @xmath141 \|x_n\|^{a+1}}{1+\delta t \|f(x_n)\|}. \label{ch5meantamed9 } \end{aligned}\ ] ] for @xmath142 , which is equivalent to @xmath143 , becomes @xmath144 from the above discussion on @xmath99 and @xmath105 , it follows that on @xmath111 , if @xmath142 and @xmath128 then we have @xmath145 taking the expectation in both sides of , using the relation @xmath146 , @xmath147 , @xmath148 and @xmath149 leads to @xmath150\mathbb{e}\|y_n\|^2 .
\end{aligned}\ ] ] iterating the last inequality leads to @xmath151^n\mathbb{e}\|x_0\|^2 .
\end{aligned}\ ] ] to have the stability of the ncts scheme , we should also have @xmath152 that is @xmath153}{2k^2+\lambda^2c^2 } , \end{aligned}\ ] ] and there exists a constant @xmath91 such that @xmath154 as in the proof of theorem [ nt1 ] , we obviously have @xmath155 . note that from the studies above , we can deduce the linear stabilities of schemes and for sdes without jump by setting @xmath156
. however , by setting @xmath157 we obtain the nonlinear stability of semi - tamed scheme without jump performed in @xcite .
the goal of this section is to provide some practical examples to sustain our theoretical results in the previous section .
we compare the stability behaviors of the tamed scheme and the semi - tamed scheme with those of numerical schemes presented in @xcite .
more precisely , we test the stability of the semi - tamed scheme , tamed scheme , backward euler and split - step backward euler schemes with different stepsizes .
we denote by @xmath158 all the approximated solutions from those schemes . here
we consider the following linear stochastic differential equation @xmath159,\quad x_0=1 , \end{aligned}\ ] ] where the intensity of the poisson process is taken to be @xmath160 .
so @xmath161 , which ensures the linear mean - square stable of the exact solution .
we can easily check from the theoretical results in the previous section that for @xmath162 the semi - tamed and the tamed euler scheme reproduce the linear mean - square property of the exact solution . in figure [ fig02 ]
, we illustrate the mean - square stability of the semi - tamed scheme , the tamed scheme , the backward and the spli - step backward euler scheme @xcite for different stepsizes .
we take @xmath163 and @xmath164 , and generate @xmath165 paths for each numerical method .
we can observe from figure [ fig02 ] that the semi - tamed scheme works well with the backward and spli - step backward euler schemes .
we can also observe that the semi - tamed scheme works better than the tamed euler scheme , and in some case overcomes the backward and spli - step backward euler schemes . for nonlinear stability , we consider the following nonlinear stochastic differential equation @xmath166,\quad \quad x_0=1.\end{aligned}\ ] ] the poisson process intensity is @xmath167 , @xmath168 and @xmath169 . we take @xmath170 and @xmath171 . indeed , we obviously have @xmath172 , @xmath173 and @xmath174 .
it follows that the exact solution is exponentially mean - square stable .
one can easily check that for @xmath175 from theoretical results the semi - tamed and the tamed euler reproduce the exponentially mean - square stability property of the exact solution .
figure [ fig03 ] illustrates the stability of the tamed scheme , the semi - tamed scheme , the backward euler and the split - step backward euler for different stepsizes .
we take @xmath176 and @xmath177 and generate @xmath178 samples for each numerical method
. we can observe that all the schemes have the same stability behavior , although tamed and semi - tamed schemes are very efficient and backward euler and the split - step backward euler less efficient as nonlinear equations are solved for each time step .
this project was supported by the robert bosch stiftung through the aims arete chair programme .
0.01 0.01 0.01 0.01 0.01 0.01
00 m. hutzenthaler , a.jentzen and p. e. kloeden .
divergence of the multilevel of the multilevel monte carlo euler euler method for nonlinear stochastic differential equations . , 23(5)(2013 ) , 19131966 .
chengming huang .
exponential mean square stability of numerical methods for systems of stochastic differential equations .
, 236 ( 2012 ) 40164026 .
higham , and p.e .
umerical methods for nonlinear stochastic differential equations with jumps . ,
101 ( 2005 ) 101119
. f. c klebaner .
introduction to stochastic calculus with applications .
, london wc2h 9he , 2004 .
x. zong , f. wu and c.huang . convergence and stability of the semi - tamed euler scheme for stochastic differential equations with non - lipschitz continous coefficients . , 228 ( 2014 ) , 240250 .
x. wang , and s. gan . compensated stochastic theta methods for stochastic differential equations with jumps . , 60(2010 ) , 877887 .
d. j. higham and p. e. kloeden .
convergence and stability of implicit methods for jump - diffusion systems . , 3(2006 ) , pp .
e. platen and n. bruti - liberati .
solution of stochastic differential equations with jumps in finance . ,
berlin , 2010 . | under non - global lipschitz condition , euler explicit method fails to converge strongly to the exact solution , while euler implicit method converges but requires much computational efforts .
tamed scheme was first introduced in @xcite to overcome this failure of the standard explicit method .
this technique is extended to sdes driven by poisson jump in @xcite where several schemes were analyzed . in this work ,
we investigate their nonlinear stability under non - global lipschitz and their linear stability . numerical simulations to sustain the theoretical results
are also provided .
stochastic differential equation , linear stability , exponential stability , jump processes , one - sided lipschitz . |
consider a short random walk of @xmath0 steps in @xmath1 dimensions where each step is of unit length and whose direction is chosen uniformly .
following @xcite , we let @xmath2 and denote by @xmath3 the probability density function of the distance @xmath4 to the origin of this random walk . this paper will be concerned with the even moments of these random walks .
define @xmath5 as the @xmath6 moment of the probability density function .
we know that for non - negative integers @xmath7 , @xmath8 is given by @xmath9 for given @xmath10 , let @xmath11 be the infinite lower triangular matrix with entries @xmath12 for row indices @xmath13 and columns entries @xmath14
. then the moments @xmath15 are given by the row sums of @xmath16 .
for a good history of these moments , and random walks in general , see @xcite . for example , the upper corner of @xmath17 and @xmath18 are given below .
@xmath19\ ] ] @xmath20\ ] ] @xmath21\ ] ] the lower triangular entries of @xmath22 are the squares of the binomial coefficients @xmath23 and those in @xmath24 are known as the naryana numbers ( * ? ? ?
* a001263 ) . using these observations about @xmath22 and @xmath24
, it is easy to observe that all of the coefficients of @xmath22 and @xmath24 are integers .
a quick glance at @xmath18 shows that this is not always true .
it was stated that @xmath25 in @xcite .
we define @xmath26 using this notation we see that @xmath27 and @xmath28 . it is not immediately that @xmath29 is well defined and finite for all @xmath10 , ( although we will show that this is the case ) . in section [ sec : a z ] we show that [ thm : main ] for @xmath30 we have @xmath31 .
this is not best possible . in section [ sec : backwards ] we prove the opposite direction [ thm : backwards ] for @xmath30 we have @xmath32 .
we conjecture that this is in fact best possible .
that is , we conjecture [ conj : main ] for @xmath30 we have @xmath33 .
we present evidence for this conjecture in section [ sec : small ] and [ sec : evidence ] .
next we consider a result by borwein , nuyens , straub and wan in @xcite about the modularity of moments .
they showed that for primes @xmath34 , we have @xmath35 we extend this in section [ sec : mod ] to get [ thm : mod ] let * @xmath36 be prime with @xmath37 , or * @xmath38 be prime with @xmath39 . then @xmath40
if @xmath41 with @xmath34 prime then @xmath42 it is worth remarking that if both @xmath43 and @xmath44 are prime with @xmath45 ( and hence @xmath46 ) , then clearly @xmath47 by the chinese remainder theorem . in section [ sec : conc ]
we discuss some of the open problems related to this research .
to prove theorem [ thm : main ] , we make use of the following remark and lemma : there are multiple equivalent ways of representing @xmath49 .
the three most common that we will use are : @xmath50 [ lem : gcd ] for integers @xmath51 we have @xmath52 let @xmath53 and @xmath54 .
let @xmath55 and @xmath56 be the products of these sequences .
let @xmath34 be a prime number and @xmath57 be the @xmath34-adic valuation of @xmath4 .
we see that for @xmath58 that there is at most one term in @xmath59 that is divisible by @xmath60 . without loss of generality
we may assume that such a term , if it exists , is in @xmath61 .
we see that @xmath62 for all @xmath7 by translation .
further , if there exists a term in @xmath63 that is divisible by @xmath60 , then , by translations we can assume that this term is divisible by an arbitrarily high power of @xmath34 .
hence we can assume that , if such a term exists , then we can find a translate of this sequence so that @xmath64 we see that if @xmath65 then there are at most @xmath66 terms in @xmath67 are are divisible by @xmath68 .
we see that if @xmath69 then there are at most @xmath70 terms in @xmath67 are are divisible by @xmath68 . by chinese remainder theorem
we can find such a @xmath71 so that both the inequalities are exact .
this gives us that @xmath72 and moreover there exists a @xmath71 so that this is exact .
we observe that @xmath73 observe that if @xmath74 then @xmath75 if @xmath76 then @xmath77
if @xmath78 then @xmath79 lastly if @xmath80 then @xmath81 hence @xmath82 which gives that @xmath83 as required .
it is worth remarking that for any fixed @xmath84 , we can find tighter lower bounds for the @xmath85 by using directly .
this can be used to tighten the results of theorem [ thm : main ] for specific @xmath10 .
unfortunately even when tightened in this way , we can not achieve the conjectured bound .
see table [ tab : compare ] .prime factorization of eq and @xmath86 . [
cols="<,<,<",options="header " , ]
we have computationally checked that for all @xmath87 that conjecture [ conj : main ] holds , further , using the techniques of theorems [ thm : nu=3 ] and [ thm : nu=4 ] we have computationally verified that for all @xmath88 and all @xmath7 that conjecture [ conj : main ] holds .
it is not unreasonable to think that conjecture [ conj : main ] can hold in general . indeed ,
if we plot the non - integer entries in the lower triangular part of @xmath11 and colour them based on the prime factorization of their denominators in reduced form we obtain the fractal pattern seen in figure ( [ frac_fig ] ) .
this suggests that there is far more structure to the matrix @xmath11 that we are currently exploiting .
we note that from equation @xmath89 combined with theorem [ thm : main ] we would be able to prove that @xmath90 .
we conjecture that @xmath91 . in this image of @xmath92 , denominators are coloured red for 2 , blue for 3 , green for 7 and orange for @xmath93 . if the denominator had contained any additional factors of @xmath94 , @xmath95 or @xmath96 then we would have coloured this value black .
none occurred . assuming that primes always give rise to the associated fractals early on , as seen in figure [ frac_fig ] , we would be led to believe that @xmath97 . ]
we rewrite ( [ moment_def ] ) as @xmath99 let @xmath36 be prime with @xmath37 or let @xmath38 be prime with @xmath39 .
we claim that there does not exist indices @xmath100 such that @xmath101 and @xmath102 .
indeed , this would lead to @xmath103 if @xmath36 then @xmath37 by assumption and hence @xmath104 , a contradiction .
if @xmath105 then @xmath39 by assumption and hence @xmath106 , a contradiction .
if instead @xmath107 and @xmath108 it is easy to see that there does not exist indices @xmath100 such that @xmath109 and @xmath110 .
we consider 2 cases : if there exists @xmath111 such that @xmath112 then clearly @xmath113 for @xmath114 and hence @xmath115 assume that @xmath116 for all @xmath111 . if @xmath36 we see that @xmath117 and @xmath118 .
we further see that at most one term in the denominator is divisible by @xmath34 .
hence @xmath119 can be written as @xmath120 where @xmath121 , and thus is equivalent to @xmath122 mod @xmath34 . if @xmath105 we see that @xmath118 .
we further see that no term in the denominator is divisible by @xmath34 .
hence @xmath119 can be written as @xmath120 where @xmath121 , and thus is equivalent to @xmath122 mod @xmath34 . if @xmath41 and @xmath108 we see that @xmath123 and @xmath124 .
we further see that we have at most @xmath125 factors of @xmath34 in the denominator , with equality only if @xmath126 for all @xmath127 .
hence @xmath119 can be written as @xmath128 where @xmath121 , and thus is equivalent to @xmath122 mod @xmath129 .
thus there are only @xmath0 terms in the sum for @xmath8 which are not @xmath130 ( resp @xmath131 ) , namely when @xmath112 for some @xmath7 . in this case
the term is @xmath132 ( resp @xmath133 ) hence @xmath134
we showed in section [ sec : small ] that conjecture [ conj : main ] held for the case @xmath135 and @xmath136 .
it is probably that this technique could be extended computationally for any fixed @xmath10 , although this is not clear .
it is not clear that this technique would be extendable to arbitrary @xmath10 without additional ideas . in section [ sec : mod ]
we showed how the ideas of modularity of @xmath137 could be extended to @xmath107 or @xmath138 .
it appears that something is also happening in the case when @xmath139 and @xmath140 , although it is unclear how one would prove this .
there are most likely many other relations that can be found when considering @xmath141 modulo a well chosen prime power .
the authors would like to thank jon borwein for many useful discussions and suggestions , without which this paper would not have been possible . | we build upon previous work on the densities of uniform random walks in higher dimensions , exploring some properties of the even moments of these densities and extending a result about their modularity . . |
our goal is to treat relativistically the response of an electron system to a periodic electric field in the framework of 1 + 1 dimensional qed ( qed@xmath1 ) .
we assume the presence of a classical periodic electric mean field , i.e. the electromagnetic field variable @xmath2 is the sum of this classical field and the quantum fluctuations .
the interaction of the fermions with this classical field is treated exactly , whereas the quantum fluctuations of the electromagnetic field are taken into account at 1-loop order . requiring that the classical field is a mean field enables us to investigate a wigner crystal ( wigner 1934 , grimes 1979 ) relativistically . to get a better understanding of such an electron system , it is unavoidable to investigate its non - perturbative response to a classical periodic electric field , that is the issue of the present work .
the classical periodic electric field is chosen in the form : @xmath3 with amplitude @xmath4 and wavenumber @xmath5 .
the homogeneous term @xmath0 corresponds to the chemical potential of the system .
in qed the expectation value of the gauge field vanishes . to take into account a mean field self - consistently in this theory we should modify its generating functional .
our method is based on defining a sector of qed for field configurations belonging to quantum fluctuations around a given vacuum expectation value @xmath6 .
we multipy the integrand of the common generating functional of qed@xmath1 by the factor 1 written in the form @xmath7 where @xmath8 is a vector in the space of the vector potential configurations and with @xmath9 the spacetime volume .
then , we find @xmath10 where the functional @xmath11 is the generating functional of the sector belonging to vector potential configurations in a hypersurface orthogonal to @xmath8 .
@xmath12 is the electromagnetic and @xmath13 is the dirac field , @xmath14 and @xmath15 are the action of the electromagnetic field and the dirac action , resp . , and @xmath16 is the gauge fixing parameter . by shifting the integration variable according to @xmath17 ,
the projected generating functional becomes @xmath18 self - consistency requires that @xmath19 this condition fixes the value of @xmath20 . to fulfill the requirement of ( [ constraint ] ) an extra renormalization condition
is needed to every single loop correction . in our treatment
the form of @xmath21 coincides with ( [ meanfield ] ) .
calculations of any physical quantity is performed by taking an expectation value of that quantity with the modified generating functional . to determine the lindhard function we should calculate the polarization @xmath22
the form of which in qed is : @xmath23 where @xmath24 is the electron propagator .
diagrammatically @xmath22 corresponds to the diagram in fig . 1 .
( without the external legs ) . in fig .
1 . the solid lines represent the electron propagator and the wavy lines correspond to the photon propagator .
since we are interested first in the response of the electron system to an arbitrary periodic potential , @xmath4 and @xmath5 should be considered as free parameters in the lindhard function .
the electron propagator belonging to the projected generating functional contains all the effects of the periodic classical field and the propagator itself looses its translational symmetry due to this sinusoidal vector potential .
another consequence of the periodic classical field is that the electron propagator from ( [ zqedp ] ) does not have analytic form , the linhard response function can only be obtained numerically .
first , we write down @xmath25 in momentum space according to : @xmath26 in momentum space the two - vector @xmath27 is the momentum of the photon in fig .
1 . coming from the left , while @xmath28 denotes the momentum of the outgoing photon . performing the integrations in ( [ pikkp ] )
we get : @xmath29 where @xmath30 and @xmath31 are positive energy eigenvalues corresponding to the positive and negative frequency eigenspinors of the dirac hamiltonian @xmath32 $ ] with the potential ( [ meanfield ] ) , and @xmath33 contains products of the eigenspinors of @xmath32 $ ] in the following form : @xmath34 @xmath35 and @xmath36 are parts of the bloch type eigenspinors ( nagy 2000 ) .
the first dirac delta in ( [ pi ] ) expresses the breaking of the translational symmetry due to the periodic field .
we expect that ( [ pi ] ) at close to @xmath37 has a maximum value for @xmath38 when the fermi surface is positioned close to the boundary of the brillouin zone .
it would mean that there is no energy exchange between the two electrons in fig . 1 .
but they exchange momentum @xmath39 .
it reminds us to the so called peierls phase transition where near the fermi surface electrons can jump from one side of the fermi surface to the other without any energy cost .
the instability caused by this procedure is stabilized by a gap in the dispersion relation arising at the edge of the fermi surface . in our treatment
the stability is ensured by the periodic electric mean field which produces a gap at the edge of the brillouin zone .
the authors would like to thank j. polonyi and g. plunien for the useful discussions .
this work was supported by the projects otka t023844/97 , daad - mb 27/1999 and nato sa ( pst.clg 975722)5066 .
numerical computations are in progress on the computer cluster donated by the alexander von humboldt foundation .
grimes , c. c. and adams , g. , phys .
rev . lett . * 42 * , 795 ( 1979 ) + nagy , s. , sailer , k. , heavy ion phys . * 11 * , nos 1 - 2 , 67 ( 2000 ) + wigner , e. p. , phys . rev . *
46 * , 1002 ( 1934 )
fig . 1 . : diagram of @xmath40 . | we give a quantum field theoretical treatment of a one dimensional electron system with a fixed chemical potential @xmath0 .
the non - perturbative lindhard response function is found for an electron system in a sinusoidal potential . |
the notion of a _ crucial word _ with respect to a given set of prohibitions is introduced in @xcite .
a word is crucial if it avoids a given set of prohibitions but any extension of the word to the right by a letter of the alphabet in question does not avoid the prohibitions .
crucial words were studied with respect to several sets of prohibitions in the literature @xcite . a crucial word is called _ bicrucial _ if it is also not possible to extend it to the left without creating a prohibition .
bicrucial words are called _ maximal words _ in the literature and they were studied with respect to so - called _ squares _ and _ abelian squares _ @xcite .
an example of a ( bi)crucial word over the alphabet @xmath12 with respect to squares / abelian squares is the _ zimin word _
@xmath13 defined recursively as @xmath14 and @xmath15 .
an extension of the notion of a crucial word to a crucial permutation @xcite was closely related to the _ theory of patterns in permutations and words _ ( see @xcite for a comprehensive introduction to the area ) .
a pattern is a permutation or word @xmath0 involving each letter from an alphabet @xmath16 for some @xmath17 , and an occurrence of @xmath0 in another permutation or word @xmath3 is a subsequence of @xmath3 that is order isomorphic to @xmath0 . for a permutation @xmath3 , its _ reduced form _
is the permutation obtained from @xmath3 by substituting the @xmath18th largest element by @xmath18 . for example , the reduced form of 2754 is 1432 .
thus , an occurrence of @xmath0 in @xmath3 is a subsequence of @xmath3 whose reduced form is @xmath0 .
when occurrences of @xmath0 are required to be formed by consecutive elements , @xmath0 is called a _
consecutive pattern_. if there are two occurrences of a consecutive pattern of length at least 2 in a permutation @xmath3 that are next to each other , we say that @xmath3 contains a square .
( bi)crucial permutations with respect to squares were studied in @xcite , while the present paper deals with _ arithmetic occurrences _ of monotone patterns in permutations and ( bi)crucial permutations associated with them ( see below for precise definitions ) .
the celebrated _ van der waerden s theorem _ states that for any given positive integers @xmath19 and @xmath17 , there is some number @xmath20 such that if the integers @xmath21 are colored , each with one of @xmath19 different colors , then there will be at least @xmath17 integers forming an arithmetic progression all of the same color .
an intriguing fact is that it is impossible to extend the van der waerden s theorem to the case of infinite permutations .
partially , it is explained by existence of so - called anti - monotone permutations in which no three elements are increasing or decreasing in positions forming an arithmetic progression ; see @xcite .
however , it is interesting to study the structure of arithmetic progressions in infinite permutations . in particular
, makarov @xcite studied _ arithmetic complexity _ , defined similarly to the notion of arithmetic complexity of words @xcite , for a number of well - known infinite permutations .
for some relevant work , see @xcite .
let @xmath1 be a permutation .
then for a fixed @xmath22 , @xmath23 is an _
arithmetic subsequence _ of length @xmath17 with difference @xmath24 , assuming @xmath25 and @xmath26 .
if an occurrence of a pattern form an arithmetic subsequence , we refer to the occurrence as an _
arithmetic occurrence _ of the pattern .
a permutation @xmath3 _ avoids arithmetically _ a pattern @xmath0 if @xmath3 contains no arithmetic occurrences of @xmath0 .
a permutation is _ @xmath5-anti - monotone _ if it avoids arithmetically the patterns @xmath6 and @xmath7 .
see @xcite for relevant work , where the operation of taking ( the usual group - theoretic ) inverse can be applied to all permutations to enter our domain . a useful fact appearing in @xcite , to be used throughout the paper , is that ( 3,3)-anti - monotone permutations exist of arbitrary length . for example , a way to construct a ( 3,3)-anti - monotone permutation of length @xmath27 is to take two ( 3,3)-anti - monotone permutations of length @xmath28 , say @xmath29 and @xmath30 and to consider the _ shuffle down _ of these permutations @xmath31 to get the sought permutation ; taking two ( 3,3)-anti - monotone permutations @xmath29 and @xmath32 would give a ( 3,3)-anti - monotone permutation of length @xmath33 , say , @xmath34 .
an _ extension _ of a permutation @xmath3 of length @xmath28 to the right ( resp . ,
left ) is a permutation @xmath35 ( resp . , @xmath36 ) of length @xmath37 such that @xmath38 and @xmath39 is obtained from @xmath3 by adding 1 to each element that is large than or equal to @xmath40 .
for example , the set of all extensions to the right of the permutation @xmath41 is @xmath42 , while the set of all extensions to the left of @xmath43 is @xmath44 .
a permutation @xmath3 is _ @xmath5-anti - monotone crucial _ ( resp .
_ @xmath5-anti - monotone bicrucial _ ) if @xmath3 is anti - monotone but any extension of @xmath3 to the right ( resp . , and to the left )
is not @xmath5-anti - monotone .
@xmath5-anti - monotone crucial permutations and @xmath5-anti - monotone bicrucial permutations will be called by us _
@xmath5-crucial _ and _ @xmath5-bicrucial _ permutations , respectively , for brevity .
for example , the permutation 216453 is @xmath45-crucial , while the permutation 73418562 is @xmath45-bicrucial .
if @xmath1 is a permutation , then its _ reverse _ , @xmath46 , is the permutation @xmath47 , while the _ complement _ of @xmath3 , @xmath48 , is obtained from @xmath3 by replacing an element @xmath18 by @xmath49 , for @xmath50 .
for example , if @xmath51 then @xmath52 and @xmath53 .
clearly , if @xmath3 is a @xmath5-(bi)crucial permutation then @xmath48 is an @xmath54-(bi)crucial permutation .
thus in our studies , without loss of generality , we can assume that @xmath55 .
another straightforward fact is that if @xmath3 is a @xmath5-bicrucial permutation then @xmath46 is an @xmath54-bicrucial permutation .
we are interested in the following questions in which we assume that @xmath8 : do there exist @xmath5-(bi)crucial permutations ?
if so , what is the minimum length of such permutations ?
do arbitrary long @xmath5-(bi)crucial permutations exist ? in this paper we will prove that arbitrary long @xmath5-crucial and @xmath5-bicrucial permutations exist for any @xmath8 .
moreover , we will show that the minimum length of a @xmath5-crucial permutation , denoted @xmath56 is @xmath57 , while the minimum length of a @xmath5-bicrucial permutation is bounded from above by @xmath58 .
@xmath5-(bi)crucial permutations of smallest length are called _ minimal _ @xmath5-bicrucial permutations .
the paper is organized as follows . in section [ kl - crucial - lower ]
we give a lower bound for @xmath56 , while an upper bound for @xmath56 is discussed in section [ kl - crucial - upper ] . in section [ arbitrarylong ]
we discuss existence of arbitrary long @xmath5-crucial permutations and in section [ kl - bicrucial ] we discuss @xmath5-bicrucial permutations . finally , in section [ final ] we state some open problems .
in this section we will prove that for @xmath8 , @xmath56 , the minimum length of a @xmath5-crucial permutation is @xmath59 .
let @xmath1 be a @xmath5-crucial permutation of minimal length , that is , @xmath60 . if a subsequence @xmath2 is an arithmetic occurrence of the pattern @xmath61 ( resp . , @xmath62 ) we say that the subsequence is an _ ap(@xmath63)-up permutation _
( resp . , _
ap(@xmath63)-down permutation _ ) .
further we let @xmath64 ( resp . ,
@xmath20 ) be the set of all ap(@xmath65)-up permutations ( resp .
, ap(@xmath66)-down permutations ) of the form @xmath67 ( resp .
, @xmath68 ) for some @xmath69 .
we have @xmath70 because extending @xmath3 to the right by @xmath37 we must get a prohibited monotone subsequence , but @xmath37 can not be the last element in a decreasing subsequence of length @xmath66 . similarly , @xmath71 because extending @xmath3 to the right by the smallest element @xmath72 we must get a prohibited monotone subsequence , but @xmath72 can not be the last element in an increasing subsequence of length @xmath17
. we let @xmath73 ( resp .
@xmath74 ) be the minimum ( resp . , maximum ) possible rightmost element in a permutation in @xmath64 ( resp .
, in @xmath20 ) .
next we will prove that @xmath75 .
suppose @xmath76 and consider extending @xmath3 to the right by an element @xmath77 .
we claim that the obtained permutation , say @xmath39 , is @xmath5-anti - monotone which contradicts to the fact that @xmath3 is a @xmath5-crucial permutation . indeed ,
if @xmath39 contains an arithmetic occurrence of the pattern @xmath6 ending with @xmath40 , then next to last element , say @xmath78 , in this occurrence must be less than @xmath73 ( the element @xmath78 of @xmath3 was unchanged in @xmath39 ) , a contradiction with the definition of @xmath73 .
on the other hand , if @xmath39 contains an arithmetic occurrence of the pattern @xmath7 ending with @xmath40 , then next to last element , say @xmath79 , in this occurrence must be larger than @xmath80 in @xmath39 and thus @xmath79 is larger than @xmath74 in @xmath3 , a contradiction with the definition of @xmath74 .
thus , we must have @xmath75 .
suppose now that an ap(@xmath65)-up permutation @xmath67 is such that @xmath81 and an ap(@xmath82)-down permutation @xmath68 is such that @xmath83 .
since @xmath75 , these permutations have no elements in common and therefore the positions of all the elements are different .
thus , we have two sets @xmath84 and @xmath85 such that @xmath69 and @xmath86 .
suppose that @xmath87 then @xmath88 ( since otherwise @xmath89 , a contradiction with @xmath86 ) .
thus @xmath90 and @xmath91 , from where the length @xmath28 of @xmath3 is no less than @xmath92 .
suppose now that @xmath93 then @xmath94 ( since otherwise @xmath95 , a contradiction with @xmath86 ) .
thus @xmath96 and @xmath97 , from where the length @xmath28 of @xmath3 is no less than @xmath98 .
summarizing our considerations , the length @xmath28 of @xmath3 can not be less than @xmath99
in this section , we consider three cases depending on whether @xmath100 , @xmath101 or @xmath102 ( see the respective subsections ) to prove that the minimum length of a @xmath5-crucial permutation is at most @xmath9 .
then summarizing considerations in these subsection and section [ kl - crucial - lower ] we will get a proof of the following theorem .
[ mainth ] for @xmath8 , the minimum length of a @xmath5-crucial permutation , @xmath56 , is equal to @xmath9 .
we begin with presenting a construction , shown schematically in figure [ mainconstruction ] , of a @xmath5-crucial permutation of any length greater than or equal to @xmath103 for @xmath104 and @xmath105 .
let @xmath3 be an arbitrary @xmath45-anti - monotone permutation of length @xmath28 , @xmath106 , shown schematically with black circles in figure [ mainconstruction ] .
rewrite the elements of @xmath3 in positions @xmath107 counted from right to form the permutation @xmath108 read from left to right , and rewrite the subsequence in positions @xmath109 counted from right to form the permutation @xmath110 read from left to right ; keep the relative order of the remaining elements of @xmath3 the same , and let them form a permutation on the set @xmath111 .
for example , let @xmath112 be a @xmath45-anti - monotone permutation and @xmath113 and @xmath114 .
then the permutation obtained by applying the construction in figure [ mainconstruction ] is 4(13)1(10)6(11)725(12)893(17)(16)(15)(14 ) .
-crucial permutations , @xmath104 and @xmath55 . ]
it is straightforward to prove that the construction is valid , namely that it presents a @xmath5-anti - monotone permutation , say @xmath39 , that is not extendable to the right . indeed , suppose @xmath39 contains an arithmetic occurrence of the pattern @xmath7 .
clearly , such an occurrence can not involve the @xmath66 rightmost elements of @xmath39 ( which are @xmath115 ) and it can involve at most one small white element ( one element from the set @xmath16 ) .
however , since only two black elements can be involved in that occurrence ( the black elements form a @xmath45-anti - monotone permutation ) and @xmath116 ( for that part , we actually need @xmath117 ) , no such occurrence is possible .
similarly , the longest arithmetic occurrence of an increasing pattern can involve at most two black elements , at most one large white element and at most one small white element , and thus such an occurrence is not possible in @xmath39 since @xmath118 . finally ,
trying to extend @xmath39 to the right by an element larger than @xmath65 , we will get an arithmetic occurrence of the pattern @xmath6 ( involving the @xmath65 smallest elements in @xmath39 ) , while extending it by an element smaller than @xmath17 , we will get an arithmetic occurrence of the pattern @xmath119 ( involving the @xmath82 rightmost elements in @xmath39 ) . applying the complement operation to the construction in figure [ mainconstruction ] and switching @xmath17 and
@xmath66 we cover the case of @xmath104 and @xmath120 . in either case ,
when @xmath100 , we have that @xmath121 . as the matter of fact , the construction in figure [ mainconstruction ] is suitable for the case @xmath122 .
the only difference is that there is one restriction on the ( 3,3)-anti - monotone permutation @xmath3 : the 6th element from the right must be smaller than the 9th element from the right in order to avoid an arithmetic occurrence with difference @xmath123 of the pattern @xmath124 involving the largest element and the smallest element .
an example of a ( 4,4)-crucial permutation obtained in this way is @xmath125 ; arbitrary long such permutations can also be constructed using an appropriate shuffling for creating ( 3,3)-anti - monotone permutations , but we do not provide more details here . note that the length of @xmath3 is @xmath126 .
-crucial permutation , @xmath117 . ] in either case , we need a construction for @xmath127-crucial permutations where @xmath116 .
we provide such a construction ( that actually works for @xmath128 as well , and gives the desired length @xmath129 ) in figure [ 4l - construction ]
. examples of permutations produced by the construction are @xmath130 for @xmath128 and @xmath131 for @xmath114 .
it is easy to see that the permutation , whose structure is presented in figure [ 4l - construction ] , is not extendable to the right .
we only need to justify that this permutation , say @xmath132 , avoids arithmetically the patterns @xmath124 and @xmath7 .
indeed , if @xmath132 contains arithmetically @xmath124 then , as it is easy to see , exactly one of the elements 1 , 2 and 3 must be involved , and exactly one of the @xmath82 largest elements must be involved , which means that the remaining middle elements in the occurrence of 1234 are either @xmath133 or @xmath134 or @xmath135 or @xmath136 ; it is straightforward to check that none of the four cases is possible .
on the other hand , to get an arithmetic occurrence of the pattern @xmath7 in @xmath132 , the difference of the arithmetic progression must be @xmath137 ( the case @xmath138 is straightforward to see not to work , while for @xmath139 we clearly will not get enough elements since the @xmath82 topmost elements can not be involved
the rightmost element can not be to the right of @xmath40 ) .
however , if @xmath137 then any arithmetic occurrence of @xmath7 will necessarily involve either @xmath140 or @xmath135 which is impossible . applying the complement operation to figure [ 4l - construction ] , we take care of the case of @xmath141-crucial permutations .
thus , in the case of @xmath101 , we have that @xmath121 .
as it was mentioned above , the permutation 216453 is @xmath45-crucial and it is of length @xmath143 .
so we only need to consider the case of @xmath144-crucial permutations with @xmath117 ( then applying the complement operation we will take care of the case of @xmath145-crucial permutations with @xmath146 ) .
-crucial permutation , @xmath117 .
] it is straightforward to see that the permutation , say @xmath3 , constructed in figure [ 3l - construction ] avoids arithmetically the pattern @xmath147 and it has no arithmetic occurrences of the pattern @xmath7 with differences @xmath138 and @xmath139 ( the latter follows from the fact that the permutation is of length @xmath148 ) .
thus @xmath137 and both of the elements around 2 must be involved in the arithmetic occurrence of @xmath7 .
if @xmath149 for some @xmath17 , then the occurrence can use at most @xmath17 elements to the right of 2 and at most @xmath65 elements to the left of 2 ( because of the two largest elements ) giving in total less than @xmath66 elements , a contradiction . on the other hand ,
if @xmath150 , the occurrence can use at most @xmath17 elements to the right of 2 , and at most @xmath17 elements to the left of 2 , thus in total less than @xmath66 elements , a contradiction .
thus , in the case of @xmath142 , we have @xmath151 .
from figure [ mainconstruction ] it is clear that arbitrary long @xmath5-crucial permutations exist of any length larger than or equal to @xmath9 for @xmath104 .
existence of arbitrary long @xmath45-crucial permutations follows , for example , from @xcite . in either case , in figure [ longcrucial ] we present two constructions of obtaining longer @xmath5-crucial permutations assuming we have a construction of shorter @xmath5-crucial permutations for any @xmath8 , which is always the case by considerations above ; however , we can not guarantee existence of respective crucial words of any length .
-crucial permutations of increasing lengths based on smaller @xmath5-crucial permutations ; @xmath8 . ] in figure [ longcrucial ] , the black circles represent schematically a given @xmath5-crucial permutation , say @xmath3 of length @xmath28 , while the white circles represent schematically an arbitrary @xmath5-anti - monotone permutation @xmath0 , respectively , of length @xmath37 and @xmath28 ( in fact , the constructions will work if we will take instead an arbitrary @xmath45-anti - monotone permutation ) .
each white element must be smaller than every black element .
as the result , permutations of lengths @xmath152 and @xmath27 , respectively , are obtained ; we use @xmath132 to denote such a permutation .
we claim that in either case , @xmath132 is also @xmath5-crucial .
indeed , non - extendability of @xmath132 to the right follows from the fact that the black elements form , in the reduced form , the @xmath5-crucial permutation @xmath3 ( instead of an arithmetic subsequence with a difference @xmath24 in @xmath3 we deal with an arithmetic subsequence with the difference @xmath153 in @xmath132 ) . on the other hand , @xmath132 is @xmath5-anti - monotone since the longest monotone arithmetic progression with an odd difference @xmath24 is of length 2 , while for even @xmath24 any arithmetic subsequence is mono - chromatic ( either entirely white or black ) and we can use the properties of @xmath3 and @xmath0 .
based on the fact that @xmath5-crucial permutations exist for any @xmath8 , we can prove that @xmath5-bicrucial permutations exist for any @xmath8 .
indeed , let @xmath154 be a @xmath5-crucial permutation of length @xmath28 , that is , @xmath154 is a @xmath5-anti - monotone permutation that is not expendable to the right . then applying reverse complement to @xmath154
, we will obtain a @xmath5-anti - monotone permutation @xmath155 that is not expendable to the left . letting @xmath154 form the largest elements in a permutation of length @xmath27 occupying every other position ( as shown schematically in figure [ longbicrucial ] by black circles ) and @xmath155 form the remain elements ( as shown schematically in figure [ longbicrucial ] by white circles ) , we obtain a @xmath5-bicrucial permutation . based on this construction , the minimal length of a @xmath5-bicrucial permutation is @xmath156 .
-bicrucial permutation of even length based on a @xmath5-crucial permutation . ] the length of a permutation obtained in figure [ longbicrucial ] is even , but to get longer @xmath5-bicrucial permutations of odd lengths , we can take any @xmath5-bicrucial permutation of even length and shuffle it with a @xmath45-anti - monotone permutation as shown schematically to the left in figure [ longcrucial ] ( one should ignore all the words in that picture ) .
it is not difficult to check that this construction works .
-bicrucial permutations , @xmath104 . ]
an alternative construction of arbitrary long @xmath5-bicrucial permutations of length @xmath157 or more for @xmath104 , is given in figure [ constrbicrucial ] ; one can notice similarity with figure [ mainconstruction ] which gave us the idea of the construction .
we note that there is some freedom in the construction in figure [ constrbicrucial ] , namely the relative order of large elements ( above the black elements ) to the right and to the left of the middle is arbitrary ; the same applies to the small elements ( below the black elements ) the relative order of those elements to the right and to the left of the middle is irrelevant .
a justification of the fact that the construction in figure [ constrbicrucial ] works is similar to our justification of figure [ mainconstruction ] : non - extendability to the right and to the left is straightforward , while any arithmetic monotone subsequences of length 4 or more can have at most one large white element and one small white element leading to the fact that a monotone subsequence of the permutation in figure [ constrbicrucial ] are of length at most 4 ( recall that the black elements form a ( 3,3)-anti - monotone permutation .
speaking of another alternative construction , which makes the length of a @xmath5-bicrucial permutation for the respective values of @xmath17 and @xmath66 smaller ( and thus is a candidate for the minimal possible such length ) , in figure [ 3l - bicrucial ] we provide a construction of @xmath144-bicrucial permutations where @xmath117 .
the length of such permutations depends on which congruence class mod 3 @xmath66 belongs to .
it is @xmath158 if @xmath159 , @xmath160 .
for example , for @xmath161 the construction gives the permutations @xmath162 , @xmath163 and @xmath164 , respectively .
-bicrucial permutations , @xmath117 . ] to justify the construction in figure [ 3l - bicrucial ] giving a permutation @xmath3 , we notice that it is not extendable to the right ( because it is based on the construction of a @xmath144-crucial permutation in figure [ 3l - construction ] ) and it is not extendable to the left ( to avoid arithmetically 123 a prospective extension must be larger than the leftmost element of @xmath3 which will lead to an arithmetic occurrence of the pattern @xmath7 with the difference @xmath123 ; to see the last claim , one has to go through three cases depending on congruence class of @xmath66 modulo 3 ) . it remains to justify that @xmath3 is @xmath144-anti - monotone .
it is easy to see that @xmath3 does not contain an arithmetic occurrence of @xmath147 . as for arithmetic occurrences of the pattern @xmath7 ,
clearly no occurrence exists for @xmath138 , and no occurrence exists for @xmath137 which follows from the fact that @xmath148 rightmost elements form a @xmath144-crucial permutation presented in figure [ 3l - construction ] , while any other occurrence is not possible because of the element 1 and three elements immediately to the right of it .
also , no arithmetic occurrence of @xmath7 is possible for the difference @xmath165 because @xmath3 does not have enough elements .
it only remains to show that no arithmetic occurrence of @xmath7 is possible for the difference @xmath123 which we do separately in each of the following three cases : * for @xmath166 , the only non - trivial case ( giving the longest possible arithmetic occurrence of the pattern @xmath167 for some @xmath63 ) is starting with the rightmost element and jumping to the left with difference @xmath123 , that is , taking from the right the elements in the 1st position , 4th , 7th , etc , @xmath168th position .
however , @xmath169 in this case , not @xmath66 , and no occurrence of @xmath7 is possible . * for @xmath170 ,
the only non - trivial case ( giving the longest possible arithmetic occurrence of the pattern @xmath167 for some @xmath63 ) is starting with the 3rd element from the right and jumping to the left with difference @xmath123 , that is , taking from the right the elements in the 3rd position , 6th , 9th , etc , @xmath171th position .
however , @xmath169 in this case , not @xmath66 , and no occurrence of @xmath7 is possible . * for @xmath172 , the only non - trivial case ( giving the longest possible arithmetic occurrence of the pattern @xmath167 for some @xmath63 ) is starting with the 3rd element from the right and jumping to the left with difference @xmath123 , that is , taking from the right the elements in the 3rd position , 6th , 9th , etc , @xmath173th position .
however , again @xmath169 in this case , not @xmath66 , and no occurrence of @xmath7 is possible .
one can check by computer that no ( 3,3)-crucial permutation exists of length 9 , and thus all ( 3,3)-crucial permutations of length 8 are ( 3,3)-bicrucial . this led us to the following question .
we have that the length of a minimal ( 3,3)-bicrucial permutation is 8 which is less than @xmath174 .
also , for @xmath117 , figure [ 3l - bicrucial ] gives examples of @xmath144-bicrucial permutations whose length is less than @xmath175 .
this led us to the following question .
finally , a rich research direction is to study ( bi)crucial permutations with respect to other sets of ( arithmetic ) prohibitions .
we believe that this direction will bring many interesting and elegant results .
a. glen , b. v. halldrsson , s. kitaev : crucial words for abelian powers , _ lect .
notes in comp .
sci . _ * 5583 * ( 2009 ) 264275 .
proceedings of the 13th conference on developments in language theory , dlt 2009 , stuttgart , germany , june 30july 3 , 2009 | a pattern @xmath0 is a permutation , and an arithmetic occurrence of @xmath0 in ( another ) permutation @xmath1 is a subsequence @xmath2 of @xmath3 that is order isomorphic to @xmath0 where the numbers @xmath4 form an arithmetic progression .
a permutation is @xmath5-crucial if it avoids arithmetically the patterns @xmath6 and @xmath7 but its extension to the right by any element does not avoid arithmetically these patterns . a @xmath5-crucial permutation that can not be extended to the left without creating an arithmetic occurrence of @xmath6 or @xmath7 is called @xmath5-bicrucial . in this paper
we prove that arbitrary long @xmath5-crucial and @xmath5-bicrucial permutations exist for any @xmath8 .
moreover , we show that the minimal length of a @xmath5-crucial permutation is @xmath9 , while the minimal length of a @xmath5-bicrucial permutation is at most @xmath10 , again for @xmath11 . +
* keywords : * crucial permutation , bicrucial permutation , monotone pattern , arithmetic pattern , minimal length |
purpose and scope of this paper is to introduce and discuss a simple hamiltonian dynamical system describing the motion of @xmath0 particles in the ( _ complex _ ) plane .
this problem is the prototype of a class of models @xcite that feature a transition from _ very simple _ ( even _ isochronous _ ) to _ _ quite complicated _ _ motions characterized by a _ sensitive dependence _ both on the initial data and the parameters ( `` coupling constants '' ) of the model .
this transition can be explained as travel on riemann surfaces .
the interest of this phenomenology illustrating the _ onset _ in a _
deterministic _ context of _ irregular _ motions is underlined by its generality @xcite , suggesting its eventual relevance to understand natural phenomena and experimental investigations .
the novelty of the model treated herein is that it allows a quite explicit mathematical treatment . here only some of our main findings are reported , without detailing their proofs : a more complete presentation will be published elsewhere @xcite .
the idea that the _ integrable _ or _ nonintegrable _ character of a dynamical system is closely related to the _ analytic _ structure of its solutions as functions of the independent variable ( `` time '' , but considered as a _ complex _ variable ) goes back to such eminent mathematicians as carl jacobi , sophia kowalewskaya , henri poincar , paul painlev and his school .
some of us heard illuminating discussions of this notion by martin kruskal , whose main ideas a synthetic if overly terse rendition of which might be the statement that _ integrability is compatible with the presence of multivaluedness but only provided this is not excessive _ can be gleaned from some papers written by himself and some of his collaborators @xcite , or by others who performed theoretical and numerical investigations motivated by his ideas @xcite .
the results presented below constitute progress along this line of thinking . for a more detailed analysis
we refer the interested reader to the papers where more complete versions are presented of our findings @xcite .
the model we introduce and discuss in this paper is characterized by the following equations of motion:@xmath1__notation _ _ : here and hereafter indices such as @xmath2 @xmath3 range from @xmath4 to @xmath0 and are defined @xmath5 superimposed dots indicate differentiations with respect to the _ real _ independent time variable @xmath6 ; the dependent variables @xmath7 _ _ _ _ are _ complex _ , and indicate the positions of @xmath0 point `` particles '' moving in the _ complex _ @xmath8-plane ; @xmath9 is the _ imaginary _ unit ; the parameter @xmath10 is _ positive _ , and it sets the time scale via the basic period@xmath11the @xmath0 quantities @xmath12 are arbitrary coupling constants , but in this paper we restrict consideration to the case in which they are all _ real _ and moreover satisfy the `` semisymmetrical '' restriction @xmath13entailing that the two particles with labels @xmath4 and @xmath14 are _ equal _ , while particle @xmath0 is _
different_. more special cases are the `` fully symmetrical '' , or``integrable '' , one characterized by the equality of _ all _ @xmath0 coupling constants,@xmath15and the `` two - body '' one , with only one nonvanishing coupling constant , @xmath16 in this latter case clearly @xmath17and the remaining _ two - body _ problem is easily solvable,@xmath18 \nonumber \\ \fl \qquad\qquad\quad -(-)^{s}\,\big\ { \frac{1}{4}\ [ z_{1}(0)-z_{2}(0 ) ] ^{2}+f\ \frac{\exp ( 2\,\rmi\,\omega \,t)-1}{2\,\rmi\,\omega } \big\ } ^{\,1/2}% \bigg ] ~ , \quad s = 1,2~.\label{twobodyc}\end{aligned}\]]the justification for labelling the fully symmetrical case ( [ integr ] ) as `` integrable '' will be clear from the following ( or see section 2.3.4.1 of @xcite ) .
the treatment of the more general case with @xmath0 different coupling constants @xmath12 is outlined in @xcite .
note that the equations of motion ( [ eqmot ] ) are of `` archimedian '' , rather than `` newtonian '' , type , inasmuch as they imply that the `` velocities '' , rather than the `` accelerations '' , are determined by the `` forces '' .
these equations of motion are hamiltonian , indeed they follow in the standard manner from the hamiltonian function @xmath19 ~. \label{h}\]]and they can be reformulated @xcite as , still hamiltonian , _ real _ ( and _ covariant _ , even _
rotation - invariant _ ) equations describing the motion of three point particles in the ( _ real _ ) horizontal plane . the following _ qualitative _ analysis ( confirmed by our _ quantitative _ findings ,
see below ) is useful to get a first idea of the nature of the motions entailed by our model . for _ large _ values of ( the modulus of ) @xmath20
the `` two - body forces '' represented by the last two terms in the right - hand side of ( [ eqmot ] ) become _ negligible _ with respect to the `` one - body ( linear ) force '' represented by the first term , hence in this regime @xmath21 entailing @xmath22 one thereby infers that , when a particle strays far away from the origin in the complex @xmath8-plane , it tends to rotate ( clockwise , with period @xmath23 ) on a circle : hence the first _ qualitative _ conclusion that _ all motions are confined_. secondly , the _ two - body _ forces cause a _
singularity _ whenever there is a _ collision _ of _ two _ ( or all _ three _ ) of the particles , and become dominant whenever _ two _ particles get very close to each other , namely in the case of _ near misses_. but if the three particles move _ aperiodically _ in a _ confined _ region ( near the origin ) of the _ complex _
@xmath8-plane , an _ infinity _ of _ near misses _ shall indeed occur . and
since the outcome of a _ near miss _ is generally quite _ different _
( whenever the two particles involved in it are _ different _ ) depending on which side the particles slide past each other and this , especially in the case of _ very close _ near misses , depends _ sensitively _ on the initial data of the trajectories under consideration we see here a mechanism causing a _ sensitive dependence _ of the time evolution on its initial data .
this suggests that our model ( [ eqmot ] ) , in spite of its simplicity , might also support quite complicated motions , possibly even displaying an `` unpredictable '' evolution in spite of its _ deterministic _ character .
this hunch is confirmed by the results reported below . to investigate the dynamics of our `` physical '' model ( [ eqmot ] )
it is convenient to introduce an `` auxiliary '' model , obtained from it via the following change of dependent and independent variables:@xmath24note that initially the coordinates @xmath20 and @xmath25 coincide:@xmath26the equations of motion of the auxiliary model follow immediately from ( eqmot ) via ( [ zita ] ) ( or , even more directly , by noting that , for @xmath27 @xmath28 and @xmath29:@xmath30here of course the appended prime denotes differentiation with respect to the ( _ complex _ ) variable @xmath31 .
the definition of @xmath32 implies that as the ( _ real _ ) time variable @xmath6 evolves onwards from @xmath33 the _ complex _
variable @xmath31 travels round and round , making a full tour ( counterclockwise ) in every time interval @xmath34 on the circle @xmath35 the diameter of which , of length @xmath36 lies on the imaginary axis in the _ complex _
@xmath31-plane , with one end at the origin , @xmath37 and the other at @xmath38 ( draw this circle ! ) .
hence these relations , ( [ zita ] ) , entail that if @xmath39 is _ holomorphic _ as a function of the _ complex _
variable @xmath31 in the closed disk @xmath40 encircled by the circle @xmath35 , the corresponding function @xmath41 is _ periodic _ in the _ real _ variable @xmath6 with period @xmath23 ( indeed _ antiperiodic _ with period @xmath42:@xmath43but it is easy to prove @xcite that the solution @xmath44 of ( [ eqzita ] ) is _ holomorphic _ ( at least ) in the circular disk @xmath45 centered at the origin of the _ complex _ @xmath31-plane and having the radius@xmath46one may therefore conclude that our physical system ( [ eqmot ] ) is _ isochronous with period _
@xmath47 see ( [ t ] ) . indeed an _
system is characterized by the property to possess one or more _ open _ sectors of its phase space , each having of course _ full dimensionality _ , such that _ all _ motions in each of them are _ completely periodic _ with the same _ fixed period _ ( the periods may be different in these different sectors of phase space , but must be fixed , i e. independent of the initial data , within each of these sectors ) : and in our case clearly ( at least ) _ all _ the motions characterized by initial data @xmath48 such that@xmath49are _ completely periodic _ with period @xmath23 , see ( [ periodic ] ) , since this inequality , implying ( via ( [ zitazero ] ) and ( [ dzero ] ) ) @xmath50 entails that @xmath51 is _ holomorphic _ ( at least ) in a disk @xmath45 that includes , in the _ complex _
@xmath31-plane , the disk @xmath40 .
this argument is a first demonstration of the usefulness of the `` trick '' ( zita ) , associating the auxiliary system ( [ eqzita ] ) to our physical system ( [ eqmot ] ) .
more generally , this relationship ( [ zita ] ) allows to infer the main characteristics of the _ time evolution _ of the solutions @xmath41 of our physical system ( [ eqmot ] ) from the _ analyticity properties _ of the corresponding solutions @xmath52 of the auxiliary system ( [ eqzita ] ) : indeed the evolution of @xmath53 as the time @xmath6 increases from the initial value @xmath54 is generally related via ( [ zita ] ) to the values taken by @xmath52 when @xmath31 rotates ( counterclockwise , with period @xmath55 on the circle @xmath35 in the _ complex _ @xmath31-plane and correspondingly @xmath51 travels on the riemann surface associated to its _
analytic _ structure as a function of the _ complex _
variable @xmath31 .
suppose for instance that the _ only _ singularities of @xmath51 in the _ finite _ part of the _ complex _ @xmath31-plane are _ square - root branch points _ , as it is indeed the case for our model ( [ eqmot ] ) at least for a range of values of the ratio of the coupling constants @xmath56 and @xmath57 see ( [ symm ] ) and below . ) , while the number and especially the locations of these singularities depend on the specific solution under consideration and their identification requires a more detailed knowledge than can be obtained by a local analysis _ la painlev_. ] then the _ isochronous _ regime corresponds to initial data such that the corresponding solution @xmath52 has _ no branch points _ inside the circle @xmath35 on the main sheet of its riemann surface ( i. e. that characterized by the initial data ) . moreover , if there is a _
( _ nonvanishing _ ) number of branch points inside the circle @xmath35 on the main sheet of the riemann surface of @xmath58 and a _ finite _ number of branch points inside the circle @xmath35 on all the sheets that are accessed by traveling on the riemann surface round and round on the circle @xmath35 , then clearly the corresponding solution @xmath53 is still a _ completely periodic _ function of the time @xmath59 but now its period is a _ finite integer multiple _ @xmath60 of the basic period @xmath34 the value of @xmath61 depending of course on the number of sheets that get visited along this travel before returning to the main sheet .
hence , in particular , whenever the _ total _ number @xmath62 of ( _ square - root _ ) branch points of the solution @xmath63 of the auxiliary problem ( [ eqzita ] ) is _ finite _ , the corresponding solution @xmath53 of our physical model ( [ eqmot ] ) is _ completely periodic _ , although possibly with a _ very large _ period ( if @xmath64 is _ very large _
) the value of which may depend , possibly quite sensitively , on the initial data . on the other hand if the number of ( _ square - root _ ) branch points possessed by the generic solution @xmath51 is _ infinite _ , and the riemann surface associated with the function @xmath51 has an _ infinite _ number of sheets ( as it can happen in our case , see below ) , then it is possible that , as @xmath31 goes round and round on the circle @xmath65 the corresponding value of @xmath52 travels on this riemann surface without ever returning to its main sheet , entailing that the time evolution of the corresponding function _
_ _ _ @xmath53 is _ aperiodic _ , and that it depends _ sensitively _ on the initial data inasmuch as these data characterize the positions of the branch points hence the structure of the riemann surface .
this terse analysis entails an important distinction among all these ( _ square - root _ ) _ _ _ _ branch points : the `` active '' branch - points are those located _ inside _ the circle @xmath35 on sheets of the riemann surface accessed when starting from the main sheet by traveling round and round on that circle , so that they do affect the subsequent sequence of sheets that get visited ; while the `` inactive '' branch points are , of course , those that fall _ outside _ the circle @xmath65 as well as those that are located _ inside _ the circle @xmath35 but on sheets of the riemann surface that do not get visited while traveling round and round on that circle ( starting from the main sheet ) and that therefore do _ not _ influence the time - evolution of the corresponding solution of our physical system ( eqmot ) .
this distinction is of course influenced by the initial data of the problem , that characterize the initial pattern of branch points ; clearly it is not just a `` local '' characteristic of each branch point depending only on its position ( for instance , _ inside _ or _ outside _ the circle @xmath66 : it depends on the overall structure of the riemann surface , for instance if there is no branch point on its main sheet that containing the point of departure of the travel round and round on the circle @xmath35 then clearly _ all _ the other branch points are _ inactive _ , irrespective of their location .
let us also emphasize that , whenever an _ active _ branch point is _ quite close _ to the circle @xmath65 it corresponds to a _ near miss _ involving _ two _ particles of our physical model ( [ eqmot ] ) , at which these two particles scatter against each other almost at right angles ( corresponding to the _ square - root _ nature of the branch point ) .
the difference between the cases in which such a branch point falls _ just inside _ respectively _ just outside _ the circle @xmath35 corresponds to a _ near miss _ in which the two particles slide past each other on one side respectively on the other ( see figure [ fig1 ] ) , and this makes a substantial difference as regards the subsequent evolution of our @xmath0-body system ( unless the two particles are equal ) .
the closer the _ near miss _ , the more significant this effect is , and the more _ sensitive _ it is on the _ initial data _ ,
a tiny change of which can move the relevant branch point from one side to the other of the circumference of the circle @xmath35 and correspondingly drastically affect the outcome of the _ near miss_. this is the mechanism that accounts for the fact that , when the initial data are in certain sectors of their phase space ( of course quite different from that characterized by the inequalities ( [ condiso ] ) ) , the resulting motion of the physical @xmath0-body problem ( [ eqmot ] ) is _ aperiodic _ , indeed nontrivially so : in such cases ( as we show below ) the _ aperiodicity _ is indeed associated with the coming into play of an _ infinite _ number of ( _ square - root _ ) branch points of the corresponding solution of the auxiliary problem ( [ eqzita ] ) and _ _ _ _ correspondingly with an _ infinite _ number of _ near _ _ misses _ experienced by the particles throughout their time evolution , this phenomenology being clearly characterized by a _ sensitive dependence _ on the initial data .
this mechanism to explain the transition from _ regular _ to _ irregular _ motions and in particular from an _ isochronous _ regime to one featuring _
unpredictable _ aspects was already discussed @xcite in the context of certain many - body models somewhat analogous to that studied herein .
but those treatments were limited to providing a _ qualitative _ analysis such as that presented above and to ascertaining its congruence with _ numerical solutions _ of these models .
the interest of the simpler model introduced and discussed herein is to allow a detailed , _ quantitative _ understanding of this phenomenology .
this is based on the following explicit solution of our model ( [ eqmot ] ) , obtained @xcite via the auxiliary problem ( [ eqzita ] ) : [ prueba ] @xmath67 ^{\,1/2}\cdot & & \nonumber \\
\quad \cdot \left\ { \left [ \check{w}\left ( t\right ) \right ] ^{\,1/2}-\left ( -\right ) ^{s}\,\left [ 12\,\mu -3\,\check{w}\left ( t\right ) \right ] ^{\,1/2}\right\ } ~,\qquad\quad s=1,2~ , & & \label{zsolution}\\ \label{zsolutionb } \fl \quad z_{3}\left ( t\right ) = z\,\exp \left ( -\rmi\,\omega \,t\right ) + \left ( \frac{% f+8\,g}{6\,\rmi\,\omega } \right ) ^{\,1/2}\,\left [ 1+\eta \,\exp \left ( -2\,\rmi\,\omega \,t\right ) \right ] ^{\,1/2}\,\left [ \check{w}\left ( t\right ) % \right ] ^{\,1/2}~.\end{aligned}\]]here the function @xmath68 is defined via the relation @xmath69 , \label{wtildew}\]]with@xmath70 = \bar{\xi% } + r\,\exp \left ( 2\,\rmi\,\omega \,t\right ) ~ , \label{ksi}\]]and @xmath71 implicitly defined by the _
parameter @xmath73 is defined in terms of the coupling constants @xmath74 and @xmath75 see ( [ symm ] ) , as follows:@xmath76and in ( 13)-([ksi ] ) the _ three _ constants @xmath77 @xmath78 and @xmath79 ( or @xmath80 are defined in terms of the @xmath0 _ _ _ _ initial data @xmath81 as follows : @xmath82 \fl\qquad\quad\quad r=\frac{3\,\left ( f+8\,g\right ) } { 2\,\rmi\,\omega \,\left [ 2% \,z_{3}(0)-z_{1}(0)-z_{2}(0)\right ] ^{\,2}}\,\left [ 1-\frac{1}{\check{w}(0)}% \right ] ^{\,\mu -1}~ , \label{parb}\\[4pt ] \fl\qquad\quad\quad \bar{\xi}=r\,\eta ~ , \label{parc}\\[4pt ] \fl\qquad\quad\quad\eta = \frac{\rmi\,\omega \,\left\ { \left [ z_{1}(0)-z_{2}(0)\right ] ^{\,2}+\left [ z_{2}(0)-z_{3}(0)\right ] ^{\,2}+\left [ z_{3}(0)-z_{1}(0)\right ] ^{\,2}\right\ } } { 3\,\left ( f+2\,g\right ) } -1~ , \label{pard}\\[4pt ] \fl\qquad\quad\quad\check{w}(0)=\frac{2\,\mu \,\left [ 2\,z_{3}(0)-z_{1}(0)-z_{2}(0)\right ] ^{\,2}}{\left [ z_{1}(0)-z_{2}(0)\right ] ^{\,2}+\left [ z_{2}(0)-z_{3}(0)% \right ] ^{\,2}+\left [ z_{3}(0)-z_{1}(0)\right ] ^{\,2}}~. \label{pare}\end{aligned}\ ] ] note that the constant @xmath83 is the initial value of the center of mass of the system , and indeed the first term in the right - hand side of the solution ( 13 ) represents the motion of the center of mass of the system : just a circular motion around the origin , with a constant velocity entailing a period @xmath23 .
since the rest of the motion is independent of the behavior of the center of mass , in the study of this model attention can be restricted without significant loss of generality to the case when the center of mass does not move , @xmath84 the nontrivial aspects of the motion are encoded in the time evolution of the function @xmath68 , see ( 13 ) and ( [ wtildew ] ) : let us emphasize in this connection that the dependent variable @xmath71 is that solution of the _ nondifferential _ equation ( [ eqwtilde ] ) uniquely identified by continuity , as the time @xmath6 unfolds , hence as the variable @xmath85 goes round and round , in the _ complex _
@xmath86-plane , on the circle @xmath87 with center @xmath88 and radius @xmath89 ( see ( [ ksi ] ) ) , from the initial datum assigned at @xmath90 , @xmath91 = w\left ( \bar{\xi}+r\right ) = \check{w}% \left ( 0\right ) ~ , \label{win}\]]see ( [ pare ] ) .
this specification of the initial value @xmath92 is relevant , because generally the _ nondifferential _ equation ( eqwtilde ) has more than a single solution , in fact possibly an _ infinity _ of solutions , see below .
it is clear from ( 13 ) that the time evolution of the solution @xmath93 of our model ( [ eqmot ] ) is _ mainly _ determined by the time evolution of the function @xmath68 . indeed , 1 .
the factor @xmath94 ^{\,1/2}$ ] displays a quite simple time evolution , _ periodic _ with period @xmath95 if @xmath96 and _ antiperiodic _ with period @xmath95 hence _ periodic _ with period @xmath23 if @xmath97 .
2 . if @xmath68 is _ periodic _ with period @xmath98 ,
its square root @xmath99 ^{\,1/2},$ ] appearing in the right - hand side of the solution formulas ( 13 ) , is clearly as well _ periodic _ with period @xmath98 or _
antiperiodic _ with period @xmath98 hence _ periodic _ with period @xmath100 depending whether the closed trajectory of @xmath68 in the complex @xmath101-plane does not or does enclose the ( branch ) point @xmath102 3 .
likewise the square root @xmath103 ^{\,1/2}$ ] ( see ( [ zsolution ] ) ) is also _ periodic _ with period @xmath98 or _
antiperiodic _ with period @xmath98 hence _ periodic _ with period @xmath100 depending whether the closed trajectory of @xmath68 in the complex @xmath104-plane does not or does enclose the ( branch ) point @xmath105 ( but note that a change of sign of this square root only entails an exchange between the two equal particles @xmath4 and @xmath14 ) .
4 . in conclusion
one sees that provided one considers particles @xmath4 and @xmath14 as _ indistinguishable _ then , if the time evolution of @xmath68 is _ periodic _ with period @xmath106 , @xmath107 the physical motion of the @xmath0 particles @xmath53 is also _ completely periodic _ either with the same period @xmath98 or with period @xmath108 provided @xmath106 is an _ integer multiple _ of @xmath95 ; 5 . finally , if the motion of @xmath68 is not periodic then clearly the functions @xmath109 are also not periodic .
hereafter we only discuss the time evolution of the function @xmath68 ; actually , as explained below , in this paper we limit our consideration to discussing the motion of a _ generic _ solution @xmath110 $ ] of the _ nondifferential _ equation ( [ eqwtilde ] ) .
moreover , we consider only _ generic _ solutions of ( [ eqmot ] ) , namely those characterized by initial data that exclude one of the following special outcomes : 1 .
@xmath111 takes , at some ( _ real _ ) time @xmath112 , the value @xmath113 entailing a _ pair collision _ of the @xmath14 equal particles occurring at this time , @xmath114 .
2 . @xmath111 takes , at some ( _ real _ ) time @xmath115 , the value @xmath116 entailing a _ pair collision _ of the different particle with one of the @xmath14 equal particles occurring at this time , @xmath117 or @xmath118 .
3 . the constant @xmath79 has unit modulus , @xmath119 , i. e. @xmath120 with @xmath121 _ real _ ( and of course defined @xmath122 ) which entails a _ triple collision _ of the @xmath0 particles occurring at the time @xmath121 , @xmath123 .
4 . @xmath111 vanishes at some ( _ real _ ) time @xmath121 , @xmath124 ( but , as our notation suggests , this case ( c ) is just a subcase of ( c ) , although this is not immediately obvious from ( zsolution ) but requires using also ( [ ksi ] ) and ( [ eqwtilde ] ) ) . the initial data that give rise to solutions having one of these singularities form a set of null measure .
it can be easily seen that these singular solutions @xmath53 of our physical problem ( [ eqmot ] ) correspond via ( [ zita ] ) to special solutions @xmath39 of our auxiliary problem ( [ eqzita ] ) possessing a _ branch point _ that sits _
exactly _ on the circle @xmath35 in the complex @xmath31-plane : more precisely , 1 . a _ square - root _ branch point featured by @xmath125 and @xmath126 but not by @xmath127 , 2 .
a _ square - root _ branch point featured by all @xmath0 functions @xmath51 , 3 . a branch point featured by all @xmath0 functions @xmath51 the nature of which depends on the parameter @xmath73 .
as mentioned above , in this paper we confine our treatment to discussing the time evolution of a _ generic _ root @xmath128 of the _ nondifferential _ equation ( [ eqwtilde ] ) with ( [ ksi ] ) , and in particular to identifying for which initial data its time evolution is _ periodic _ , and in such a case what the period is .
remarkably we find out that , for ( _ arbitrarily _ ) given initial data , _ all _ these roots have at most three different periods ( one of which might be _ infinite _ , signifying an _ aperiodic _ motion ) ; periods which we are able to determine explicitly ( although the relevant formulas have some nontrivial , even chaotic " , aspects , in a sense that is made explicit below ) .
the question of identifying , among _ all _ the roots @xmath129 of this _ nondifferential _ equation ( [ eqwtilde ] ) , the `` physical '' one @xmath68 i. e. the one that evolves from the initial datum ( [ pare ] ) , and in particular of specifying the character of its time evolution among the ( at most @xmath0 ) alternatives discussed below , is a technically demanding job the solution of which shall be reported in cgss .
let us re - emphasize that the time evolution of @xmath130 $ ] coincides with the evolution of a _ generic _ root @xmath131 of ( [ eqwtilde ] ) as the independent variable @xmath132 travels ( making a complete counterclockwise tour in the _ complex _ @xmath86-plane in every time interval @xmath95 ) on the circle @xmath87 with center @xmath133 and radius @xmath89 , see ( [ ksi ] ) , and correspondingly the dependent variable @xmath71 travels on its riemann surface .
note that this riemann surface is completely defined by the single parameter @xmath73 , see ( [ mu ] ) and ( [ eqwtilde ] ) , while the circle @xmath87 is defined by the initial data of the problem , see ( [ ksi ] ) with ( 18 ) .
what therefore remains to be discussed is the _ analytic _ structure of the multivalued function @xmath131 defined implicitly by the _ nondifferential _
equation ( [ eqwtilde ] ) or , equivalently but more directly , the time dependence of the corresponding function @xmath134 $ ] . to begin with , we consider the case in which the parameter @xmath73 is _ rational _ , @xmath135with @xmath136 and @xmath64 coprime integers and @xmath64 _ positive _ , @xmath137 .
the extension of the results to the case of _ irrational _
@xmath73 is made subsequently ; although , to avoid repetitions , we present below some results in a manner already appropriate to include also the more general case with @xmath138 _
real_. in the _ rational _ case ( [ mupq ] ) the _ nondifferential _ equation that determines the `` dependent variable '' @xmath139 in terms of the `` independent variable '' @xmath6 becomes _ polynomial _ , and takes one of the following @xmath0 forms depending on the value of the parameter @xmath140 see ( [ mupq ] ) : @xmath141 ^{\,q}\,\tilde{w}^{\,p},\qquad&\mbox{if } \mu > 1 , \label{poleqa}\\ \left [ \bar{\xi}+r\,\exp \left ( 2\,\rmi\,\omega \,t\right ) \right ] ^{\,q}\,\left ( \tilde{w}-1\right ) ^{\,q - p}\,\tilde{w}^{\,p}=1,\qquad&\mbox{if } % 0<\mu < 1 , \label{poleqb}\\ \left [ \bar{\xi}+r\,\exp \left ( 2\,\rmi\,\omega \,t\right ) \right ] ^{\,q}\,\left ( \tilde{w}-1\right ) ^{\,q+\left\vert p\right\vert } = \tilde{w}% ^{\,\left\vert p\right\vert } , \qquad&\mbox{if } \mu < 0 .
\label{poleqc}\end{aligned}\]]the above expressions are polynomials ( in the dependent variable @xmath142 ) of degree @xmath143 : @xmath144 as for the boundaries of these @xmath0 cases , let us recall that @xmath145 corresponds , via ( [ mu ] ) , to @xmath146 namely , see ( [ symm ] ) , to the trivially solvable _ two - body _ case , see ( 5 ) , while @xmath147 respectively @xmath148 correspond , via ( [ mu ] ) , to @xmath149 respectively to @xmath150 and require a separate treatment , for which the interested reader is referred to cgss .
clearly the third case ( @xmath151 ) becomes identical to the first ( @xmath152 ) via the replacement @xmath153 without modifying @xmath64 ; therefore in the following , without loss of generality , we often forsake a separate discussion of this third case .
clearly the factor @xmath154 ^{\,q},$ ] which carries all the time dependence in these polynomial equations , is _ periodic _ in @xmath6 with period @xmath95 , see ( [ t ] ) ( except for the _ special _ initial conditions entailing @xmath155 in which case this factor is instead periodic with the shorter period @xmath156 ) ( for simplicity we continue to pursue our policy to consider only the _ generic _ case when this does _ not _ happen , referring the interested reader to @xcite for a more complete treatment ) . at issue
is the behaviour of the @xmath143 roots @xmath157 of our polynomial equation ( 21 ) whose coefficients evolve in time periodically with period @xmath95 .
let us indicate with @xmath158 the ( _ unordered _ ) set of these @xmath143 roots .
obviously @xmath159 is _ periodic _ with period @xmath95 , @xmath160 : after one period @xmath95 the polynomial equation is unchanged , hence the set of its @xmath143 roots is as well unchanged . but that does _ not _ imply that if one follows the time evolution of these @xmath143 roots , each of them will return to its own initial value after one period , @xmath161 @xmath162 this outcome will indeed obtain for the _ open _ domain of initial data of our problem that corresponds to the basic _ isochronous _ regime , see ( [ periodic ] ) ; but it does not happen for other initial data , in which cases for instance a _ generic _ root , say @xmath163 may after one period land in the initial position of a different root , say @xmath164 and then after one more period end up in the initial position of yet another root , @xmath165 and so on . eventually , of course , after a time @xmath166 which is a _ finite integer _
multiple @xmath167 of the basic period @xmath95 , with @xmath168 the _ generic _ root @xmath169 shall necessarily return to its initial position , @xmath170 entailing that its evolution as a function of the time @xmath6 is _ periodic _ with this period @xmath171 , @xmath172 this discussion clearly implies ( via ( [ zsolution ] ) ) that , in the case now under consideration ( with a
_ rational _ value of the parameter @xmath138 , see ( [ mupq ] ) ) , _ all _ solutions of our physical problem ( [ eqmot ] ) with ( [ symm ] ) are _ completely periodic _ with a period which is either @xmath173and @xmath143 defined by ( [ jeq ] ) , or it is @xmath100 ( see the discussion above following ( [ win ] ) ) .
the remaining , crucial question is : how does the value of the integer @xmath174 ( which might be _ quite large _ if @xmath143 is _ quite large _ ) depend on the initial data of our problem ? in this paper we outline how to calculate , for given initial data , _ all _ the possible periods of the @xmath143 roots @xmath129 of ( 21 ) , and we display formulas providing ( at most ) @xmath0 alternative values for these periods ; as already mentioned above , the explanation of how to identify which one of these @xmath0 periods corresponds to that of the `` physical '' root @xmath68 entails a more detailed treatment which is reported in @xcite . but before doing so let us emphasize that via this discussion the time evolution of our original @xmath0-body problem describing the time evolution of the three points @xmath53 in the _ complex _
@xmath8-plane has been related to the time evolution of the @xmath143 roots @xmath157 of ( 21 ) in the _ complex _
@xmath175-plane , and in particular to the way they get permuted among themselves over the time evolution after each period @xmath95 .
as we will explain below , the possible complications of the motions of our physical model ( [ eqmot ] ) are thereby related to the mechanisms at play to permute these roots among themselves when one watches their time evolution at subsequent intervals @xmath176and so on . and , in this context , it is significant to note that , whenever @xmath73 is _ irrational _ , one is in fact dealing with the dynamics of an _ infinite _ number of roots .
this suggests that whenever the number @xmath143 of roots is large , and even more so when @xmath73 is irrational ( entailing @xmath177 ) , the time evolution of our physical model ( [ eqmot ] ) might be quite complicated , perhaps calling into play the theoretical tools of statistical mechanics rather than those of few - body dynamics ( but we postpone such excursions to future publications ) .
as entailed by our discussion above , the issue of determining the value of the integer @xmath167 is tantamount to understanding the structure of the riemann surface associated with the function @xmath71 of the _ complex _
variable @xmath86 defined by the _ nondifferential _ equation ( eqwtilde ) , whose different sheets correspond of course to the different roots of this _ nondifferential _ equation . in the _ rational _ case ( [ mupq ] ) this equation is in fact _ polynomial _ of degree @xmath143 in the dependent variable @xmath175 .
specifically , what must be ascertained is the number of sheets of this riemann surface that are accessed by @xmath71 when the independent variable @xmath86 travels in the _ complex _
@xmath178-plane round and round on the circle @xmath87 , whose center @xmath88 and radius @xmath89 depend on the initial data of our physical problem , see ( [ ksi ] ) and ( 18 ) .
to this end one must gain and use a detailed understanding of the structure of this riemann surface .
again , we refer for the details of this analysis to cgss
. suffice here to report the following basic information on the _ analytic _ structure of the function @xmath179 referring to the general case with _
@xmath73 ( _ rational _ or _ irrational _ ) . the nondifferential equation ( [ eqwtilde ] ) defines a @xmath143-sheeted covering of the _ complex _ @xmath86-plane of genus _ zero _ ( of course @xmath180 is @xmath181 is irrational ) . the function @xmath182 defined implicitly by this equation features _ square - root _ branch points @xmath183 located on a circle @xmath184 centered at the origin of the _ complex _ @xmath86-plane : @xmath185@xmath186 ~,~\,~k=1,2, ... ,q~ , \label{singulb}\]]@xmath187 in the last , ( [ singulc ] ) , of these formulas it is understood for definiteness that the _ principal _ determination is taken of the @xmath73-th power appearing in the right - hand side .
the first of these formulas , ( [ singula ] ) , shows clearly that the number of these branch points is _ infinite _ if the parameter @xmath73 is _ irrational _ , and that they then sit densely on the circle @xmath184 in the complex @xmath86-plane centered at the origin and having radius @xmath188 , see ( singulc ) .
note that this entails that the _ generic _ point on the circle @xmath184 is _ not _ a branch point ( just as a _ generic _ _ real _ number is _ not rational _ ) ; but every generic point on the circle @xmath184 has , if @xmath73 is _ irrational _ , some branch point ( in fact , an _ infinity _ of branch points ! ) _ arbitrarily _ close to it ( just as every generic _ real _ number has an _ infinity _ of _ rational _ numbers _ arbitrarily _ close to it ) .
it is also important to realize that these branch points are generally on _ different _ sheets of the riemann surface associated with the function @xmath189 : hence , they are dense if one considers the circle @xmath184 in the _ complex _
@xmath86-plane , but they are not dense if one considers these branch points on the riemann surface itself . as for the second of these formulas , ( [ singulb ] ) , it is instead appropriate to the case in which the parameter @xmath73 is _ rational _ , see ( [ mupq ] ) , in which case the branch points sit again on the circle @xmath184 in the complex @xmath132-plane , but there are only a _
finite _ number , @xmath64 , of them ( and note that the factor @xmath136 appearing in the argument of the exponential in the right - hand side of this formula , ( [ singulb ] ) , is only relevant to characterize how the branch points @xmath190 are labeled via the index @xmath191 ) . both in the _ irrational _ and in the _ rational _ case at _ all _ these branch points the _ nondifferential _ equation ( [ eqwtilde ] ) has a _
double _ root which takes the _ same _ value @xmath192 .
note that this entails that circling around such a branch point in the complex @xmath86-plane corresponds to permuting @xmath14 of the @xmath143 roots @xmath193 among themselves .
in addition , the function @xmath71 possesses branch points at @xmath194 and at @xmath195 , the order of which depends on the value of @xmath73 , and is _ rational _ if the number @xmath73 is _ rational _ ( see below ) . the branch point at @xmath195 has , if @xmath196 exponent @xmath197 ( @xmath198 in the _ rational _ case ) , @xmath199while it has instead two different exponents if @xmath200 : 1 . the exponent @xmath201 ( @xmath198 in the _ rational _ case ) as given by the preceding formula ( entailing @xmath202 2 .
the exponent @xmath203 ( @xmath204 in the _ rational _ case ) as given by the following formula ( entailing @xmath205),@xmath206 the branch point at @xmath194 is , if @xmath196 of exponent @xmath207 ( @xmath208 in the _ rational _ case ) , @xmath209(note the formal analogy of this formula with the previous one , ( branchinfb ) ) , and it is instead _ absent _ if @xmath210 so that in this second case the _ only _ branch points in the finite part of the _ complex _
@xmath86-plane are those of _
square - root _ type , see ( [ singula ] ) .
this is the main cause of the difference between the results , see below , for this case ( @xmath200 ) from those for the other two cases ( @xmath152 and @xmath211 , which are on the other hand essentially equivalent among each other being related by the transformation @xmath212 @xmath213 see ( [ eqwtilde ] ) ( so that without loss of generality we often forsake an explicit discussion of the case with @xmath151 ) .
note that these results entail that , in the _ rational _ case , see ( mupq ) , making a circle around the branch point at @xmath194 , in the @xmath214 ( i. e. @xmath215 ) case when this branch point is present , causes a cyclic permutation of @xmath216 of the @xmath136 roots @xmath217 : this is particularly evident if one imagines to travel full circle around the branch point at @xmath194 in its immediate vicinity , since for @xmath218 the @xmath136 roots @xmath217 of ( [ eqwtilde ] ) are clearly divided into _ two _ sets , a first set of @xmath219 roots , disposed equispaced on a circle of small radius ( @xmath220 centered at @xmath221 in the _ complex _
@xmath175-plane , which then undergo a cyclic permutation among themselves , and a second set of @xmath64 roots , disposed equispaced on a circle in the _ complex _
@xmath175-plane centered at the origin and having a large radius ( @xmath222 ) , each of which after the operation returns instead to its original position .
the permutation experienced by the @xmath143 roots @xmath157 due to a sequence of pairwise exchanges of roots which take place whenever _ square - root _
branch points are included _ inside _ the circle @xmath87 traveled by the point @xmath223 causes a reshuffling of the roots which is nontrivial inasmuch as it depends on how many and on which pairs of roots get sequentially exchanged over each period , as determined by the number and identity of _ square - root _
branch points enclosed inside the circle @xmath224 traveled by @xmath223 and by the detailed structure of the riemann surface associated with these branch points , in particular on which sheets of this riemann surface the relevant branch points are located .
the reshuffling encompasses more roots when a second mechanism is simultaneously at play , i. e. that producing a cyclic permutation of @xmath216 roots ( especially , of course , when @xmath216 is large ) over each period , as caused by the presence of the branch point at @xmath194 : this second mechanism exists only if @xmath215 ( or @xmath225 entailing the exchange @xmath226 ) , and provided the circle @xmath87 traveled by @xmath223 _ does include _ the point @xmath227 this phenomenology causes the possible periods @xmath166 of the time evolution of the _ generic _ root @xmath228 to depend on the initial data , but remarkably we will see that , for given initial data , there are ( at most ) only @xmath0 possible values of these periods , and that they can be given explicitly in terms of the initial data and of the two numbers @xmath136 and @xmath229 see ( [ mupq ] ) .
indeed we show below that analogous results can as well be given in the case when @xmath73 is _ irrational _ , in spite of the fact that the dependence on the initial data may then be quite _
sensitive_. the correspondence of the analysis , given here in terms of the dynamics of the roots @xmath157 , with that in terms of travel on the riemann surface made above ( see paragraphs after ( [ condiso ] ) ) , including the distinction made there about _ active _ and _ inactive _ branch points , should be noted : the _ active _ branch points are those that cause a reshuffling of roots that involves the `` physical '' root @xmath68 , the _ inactive _ ones are those that do _ not _ cause a reshuffling of roots that involves the `` physical '' root @xmath68 , either because they cause no reshuffling at all being located _ outside _ the relevant circle ( @xmath87 in the _ complex _ @xmath86-plane in the context of the present analysis , @xmath35 in the _ complex _ @xmath31-plane in the context of the discussion made above when the distinction among _ active _ and _ inactive _ branch points was first introduced ) , or because they cause a reshuffling which however does not involve _ _ _ _ the physical root @xmath68 .
we now report our findings concerning the time evolution of a generic root @xmath169 , referring at first mainly to the _ rational _ case but including immediately results for the _ irrational _ case whenever it is convenient to do so in order to shorten our presentation . the remaining information on the _ irrational _ case is provided below ( see proposition 5 ) .
first of all it is useful to visualize the two circles @xmath184 and @xmath87 in the _ complex _
@xmath86-plane ( draw them ! ) : recall that the circle @xmath184 on which the branch points sit is centered at the origin and its radius @xmath230 only depends on the parameter @xmath140 see ( singulc ) , while both the center @xmath88 and the radius @xmath231 of the circle @xmath87 traveled upon by @xmath232 see ( ksi ) , _ do depend _ on the initial data , see ( 18 )
. * proposition 1 . *
if the circle @xmath87 is _ inside _ the circle @xmath233 ( i. e. @xmath234 see ( 18 ) and ( [ singulc ] ) ) , and ( a ) @xmath73 is _ inside _ the interval @xmath200 or ( b ) @xmath73 is _ outside _ this interval ( @xmath235 or @xmath211 and moreover the circle @xmath87 does _ not _ include the origin @xmath194 ( i. e. @xmath236 or equivalently @xmath237 ) , then @xmath238 i. e. the _ generic _ solution @xmath128 is periodic with period @xmath34 @xmath239 .
this outcome applies equally if @xmath73 is _ rational _ or _
irrational_. @xmath240 * remark 1*. in _ all _ the cases identified in this proposition 1 there are _ no branch points at all _ inside the circle @xmath224 : indeed the outcome detailed by this proposition 1 applies in _ all _ the cases in which this happens ( see our discussion above ) , even ( in the case with _ rational _
@xmath241 if the circles @xmath184 and @xmath87 do cross each other marginally ; and of course in _ all _ these cases _ all _ the roots @xmath157 are _ periodic _ with period @xmath242 and in the context of the physical problem ( [ eqmot ] ) the solution is characterized by the simple periodicity rule ( [ periodic ] ) .
the restriction ( [ condiso ] ) on the initial data is _ sufficient _ ( but of course not _ necessary _ ) to guarantee that we are in this regime.@xmath243 * proposition 2*. if the circle @xmath87 is _ inside _ the circle @xmath233 ( i. e. @xmath244 ) , the circle @xmath87 _ does _ include the origin @xmath194 ( i. e. @xmath245 or equivalently @xmath246 ) , and @xmath73 is _ outside _ the interval @xmath200 then in the _ rational _ case , see ( [ mupq ] ) , 1 . @xmath247 or @xmath248
if @xmath249 2 .
@xmath247 or @xmath250 if @xmath151 . in the _ irrational case _ the time evolution of the _ generic _
root @xmath128 is either _ periodic _ with the basic period @xmath34 or _ quasiperiodic _ , involving in particular a ( nonlinear ) superposition of _ two _ periodic evolutions with _ two _ noncongruent periods , specifically ( a ) with period @xmath95 and @xmath251 if @xmath196 ( b ) with period @xmath95 and @xmath252 if @xmath151.@xmath240 * remark 2*. in the case identified in this proposition 2 the only branch point _ inside _
@xmath87 is that at @xmath253 which is indeed only present if @xmath215 or @xmath254 hence in the _ rational _ case with @xmath214 ( i. e. @xmath255 , @xmath216 roots @xmath157 get cyclically exchanged among themselves , entailing that the time evolution of each of them has period @xmath256 while the remaining @xmath64 roots have period @xmath257 with an analogous phenomenology in the @xmath151 . likewise , when @xmath73 is _ irrational , _ the periodicity of the time evolution of the _ generic _ root @xmath128 has period @xmath95 if the branch point at @xmath194 is _ inactive _ ( i. e. , it does not appear on the sheet on which @xmath258 lives ) , otherwise its time evolution can be inferred by replacing @xmath86 with @xmath232 see ( [ ksi ] ) , in the formula characterizing the branch point at @xmath194 , see ( [ branchzero ] ) .
note however that for the _ special _ initial data such that the two circles @xmath184 and @xmath87 are _ concentric _
( i. e. , @xmath259 ) the _ quasiperiodic _ time evolution of @xmath128 is instead _ periodic _ ( a ) with period @xmath251 if @xmath196 ( b ) with period @xmath260 if @xmath151.@xmath240 * proposition 3*. if the circle @xmath87 is _ outside _ the circle @xmath184 ( i. e. @xmath261 ) then in the _ rational _ case , see ( [ mupq ] ) , 1 .
@xmath262 if @xmath249 , 2 .
@xmath263 if @xmath264 3 .
@xmath265 or @xmath266 if @xmath200 . in the _ irrational case
_ the time evolution of the _ generic _ root @xmath128 is _ quasiperiodic _ , involving a ( nonlinear ) superposition of _ two _ periodic evolutions with _ two _ noncongruent periods , specifically 1 . with periods @xmath95 and @xmath267 if @xmath196 2 . with periods @xmath95 and @xmath268 if @xmath225 3 . with periods @xmath95 and @xmath267 or @xmath269 and @xmath270 if @xmath200.@xmath240 * remark 3 . * in all the cases encompassed in this proposition 3 the branch points of @xmath131 in the _ finite _ part of the _ complex _
@xmath86-plane are _ all inside _ the circle @xmath87 , hence the dynamics of the roots @xmath157 can be understood in terms of the branch point of @xmath131 at @xmath195 . therefore in the _ rational _ case with @xmath214 ( i. e. @xmath152 ) _ all _ the @xmath136 roots get cyclically exchanged , so that each of them gets back to its original value after a period @xmath271 ; likewise if @xmath272 ( i. e. @xmath151 ) _ all _ the @xmath273 roots get cyclically exchanged , so that each of them gets back to its original value after a period @xmath274 in the other _ rational _ case , @xmath275 ( i. e. @xmath200 ) , @xmath136 roots gets cyclically exchanged among themselves , and the remaining @xmath276 roots get cyclically exchanged among themselves , so that the _ generic _ root @xmath169 has period @xmath271 if it belongs to the first set , and @xmath277 if it belongs to the second . and the outcome in the _ irrational _ case can as well be understood in terms of the exponent of the branch point at @xmath278 see ( [ branchinfa])-([branchinfb]).@xmath240 the situation is less straightforward ( hence more interesting ) if the two circles @xmath184 and @xmath87 _ do intersect _ each other ( i. e. @xmath279 ) .
then the parameter that plays a crucial role is the number @xmath280 of _ square - root _
branch points , sitting on the circle @xmath281 that fall _ inside _ the circle @xmath87 .
this number @xmath280 is _ finite _ , @xmath282 ( note that the case @xmath283 is taken care of by proposition 1 ) only in the _ rational _ case , to which we restrict consideration in the following proposition 4 ( and we exclude from consideration the _ nongeneric _ case in which the circle @xmath87 hits one of the branch points sitting on the circle @xmath184 )
. then _ two _ or _ three _ ( but no more ! ) different alternatives are possible for the value of the _ positive integer _ @xmath284 as detailed below .
hereafter the notation @xmath285 denotes the _ floor _ of the _ real _ number @xmath286 namely the _
largest integer number not larger than _ @xmath287 ( hence for instance @xmath288 @xmath289 .
* proposition 4*. ( i ) .
if @xmath214 ( i. e. @xmath196 see ( [ mupq ] ) ) and the origin @xmath194 is _ outside _ the circle @xmath87 ( i. e. @xmath290 ) , then @xmath167 can take one of the following @xmath0 values : @xmath291 ( ii ) .
if @xmath214 ( i. e. @xmath196 see ( [ mupq ] ) ) and the origin @xmath194 is _ inside _ the circle @xmath87 ( i. e. @xmath246 ) , then @xmath167 can take one of the following @xmath14 values:@xmath292 ( i ) .
if @xmath293 ( i. e. @xmath225 see ( [ mupq ] ) ) and the origin @xmath194 is _ outside _ the circle @xmath87 ( i. e. @xmath294 ) , then @xmath167 can take one of the following @xmath0 values : @xmath295 ( ii ) .
if @xmath293 ( i. e. @xmath225 see ( [ mupq ] ) ) and the origin @xmath194 is _ inside _ the circle @xmath87 ( i. e. @xmath246 ) , then @xmath167 can take one of the following @xmath14 values:@xmath296 ( iii ) .
the situation is more intriguing if @xmath275 ( i. e. , @xmath210 see ( [ mupq ] ) )
. then one must introduce the _ simple continued fraction _ expansion of the number @xmath297the @xmath191th _ convergent _ @xmath298 of this _ continued fraction _ expansion ( [ confuna ] ) , and in particular its numerator @xmath299 and denominator @xmath300,@xmath301are then defined by the recursions@xmath302note that these formulas apply equally if @xmath73 is _ rational _ or _ irrational _ ; of course depending whether @xmath73 is _ rational _ or _ irrational _ the _ continued fraction _ expansion does or does not terminate . in the _ rational _ case under present discussion we also introduce another sequence of _ nonnegative integers _ : given @xmath280 , let the integer @xmath304 and the period @xmath305 be defined by the following formulas : @xmath306 @xmath307 then the roots of the polynomial can have only one of the following three periods @xmath308 : @xmath309 this is the generic case ; there are however some cases in which the roots have only the two periods @xmath305 and @xmath310 .
this happens of course when @xmath311 and whenever @xmath280 takes the following special values : @xmath312 * remark 4 .
* in case ( i ) of proposition 4 the mechanism that yields periods longer than unity is the coming into play of the @xmath280 _ square - root _ branch points enclosed _ inside _ the circle @xmath87 , which cause a certain number of roots @xmath157 to exchange pairwise their roles through the time evolution . but
this phenomenology only affects some roots ; others remain unaffected , hence their periods remain unity , and this explains the first entry in ( [ jtilaa ] ) .
the precise form of the other entries in this formula , ( [ jtilaa ] ) , requires of course a more detailed treatment , see @xcite ; the outcome there depends on how many pair exchanges actually do take place , or , equivalently , how many sheets of the riemann surface get actually visited , and this depends in a fairly detailed manner on the specific structure of this surface . but note that only _ two _ different periods may emerge , differing by only _
one _ unit . in case ( ii ) of proposition 4
the second mechanism , associated with the presence of the branch point at @xmath253 comes additionally into play , causing the connection of _ all _ the @xmath280 sheets containing the @xmath280 _ square - root _
branch points , both among themselves and with the @xmath313 sheets containing the branch point at @xmath194 .
the corresponding @xmath314 roots get permuted among themselves through the time evolution , with the period indicated by the second entry in ( [ jtilab ] ) ( note that whenever @xmath136 is _ quite large _ and @xmath280 is _ close _ to its _ maximal _ value @xmath315 the resulting period is _
quite large)_. the remaining @xmath316 sheets , corresponding to the @xmath317 branch points lying on the circle @xmath184 _ outside _ the circle @xmath87 , are isolated , hence the corresponding roots do _ not _ take part in the quadrille , so that their period remains unity , as indicated by the first entry in ( [ jtilab ] ) .
the cases ( i ) and ( ii ) of proposition 4 require no additional discussion . in case ( iii ) of proposition 4 ( with @xmath275 ,
i. e. @xmath200 ) there is no branch point at @xmath194 ; hence the mechanism is now _ absent _ that previously caused the connection of all the sheets associated with the @xmath280 _ square - root _
branch points sitting on the circle @xmath184 _ inside _ the circle @xmath87 .
this implies , see @xcite , that each sheet of the riemann surface contains _ only two _ branch points , whose projections on the circle @xmath184 are separated by @xmath318 other branch points .
note that this entails that two branch points which are _ adjacent _ on the circle @xmath184 in the _ complex _
@xmath86-plane are instead topologically far apart on the riemann surface of the function @xmath319 living on sheets which are not directly connected . as a consequence
the period of the time evolution of the root @xmath128 does not change , as the initial data change causing the circle @xmath87 to change its position and dimension so that the number @xmath280 of _ square - root _ branch points enclosed in it increases , until some crucial _ square - root _
branch point gets thereby included inside the circle @xmath320 causing the connection of two separate groups of connected sheets ; and this mechanism occurs more and more frequently as @xmath280 increases more and more .
this explained _ qualitatively _ the piecewise constant behavior of the period as @xmath280 increases , characterized by shorter and shorter steps and by bigger and bigger jumps , see ( [ t2])-([jtilc ] ) .
the exact treatment of this mechanism , yielding ( [ t2])-([jtilc ] ) , is rather complicated , as indicated by the role played by the _ continued fraction _ expansion of the number @xmath321 .
the details are given in @xcite .
here we limit ourselves to emphasizing that , for given initial data , the _ generic _ root @xmath128 can have only @xmath0 possible periods , the third of which is just the sum of the first two , see ( [ jtilc]).@xmath240 let us now discuss the case in which @xmath73 is an _ irrational _ number , recalling that we are now considering initial data ( @xmath322 ) such that the two circles @xmath184 and @xmath87 _ do intersect _ each other ( the results for the other cases have been given in propositions @xmath4@xmath0 . ) .
one must then introduce the ratio @xmath323 of the length of the arc of the circle @xmath184 that is _ inside _ the circle @xmath320 to the length of the _ entire _ circle @xmath184 : of course @xmath324 expressing this ratio in terms of the initial data is an exercise in elementary plane geometry yielding the formula@xmath325where the determination of the @xmath326 function must be chosen so that @xmath327 the results for the periods are then given by the following * proposition 5*. ( i ) if @xmath215 and the origin @xmath194 is _ outside _ the circle @xmath87 ( i. e. @xmath290 ) , the time evolution of the _ generic _ root @xmath128 is still _ periodic _ with period @xmath328 @xmath329 and @xmath167 can take one of the following @xmath0 values : @xmath330note that the second and third entry only differ by one unit and moreover that , if @xmath331 the floor functions vanish , hence for _ all _ values of @xmath73 larger than @xmath14 ( and @xmath290 ) the only possible values for @xmath167 are @xmath4 or @xmath332 \(ii ) if @xmath215 and the origin @xmath194 is _ inside _ the circle @xmath87 ( i. e. @xmath246 ) , the time evolution of the _ generic _ root @xmath128 is either _ periodic _ with period @xmath95 or
_ aperiodic_. ( i ) .
if @xmath151 and the origin @xmath194 is _ outside _ the circle @xmath224 ( i. e. @xmath290 ) , the time evolution of the _ generic _ root @xmath128 is still _ periodic _ with period @xmath333 @xmath329 and @xmath167 can take one of the following @xmath0 values : @xmath334again the second and third entry only differ by one unit and moreover if @xmath335 the floor functions vanish , hence we conclude that for _ all _ values of @xmath73 smaller than @xmath336 ( and @xmath294 ) the only possible values for @xmath167 are @xmath4 or @xmath337 ( ii ) if @xmath151 and the origin @xmath194 is _ inside _ the circle @xmath338 ( i. e. @xmath246 ) , the time evolution of the _ generic _ root @xmath128 is either _ periodic _ with period @xmath95 or _ aperiodic_. 1 .
the integers @xmath136 and @xmath64 diverge with their ratio @xmath347 fixed ( see ( [ mupq ] ) ) , and 2 .
the number @xmath280 of _ square - root _
branch points _ inside _ the circle @xmath320 as well as the _ total _ number @xmath64 of _ square - root _
branch points , also diverge with their ratio fixed ( recall that _ all _ these _ square - root _
branch points sit , densely equispaced , on the circle @xmath184 in the _ complex _
@xmath86-plane although on _ different _ sheets of the riemann surface of the function @xmath348 hence this ratio coincides with the quantity @xmath323 defined , and evaluated in terms of the initial data , above , see ( [ nu ] ) ) .
this also suggests obvious extensions to the present case with _ irrational _ @xmath73 of comments contained in the remark 4 , which will not be repeated here .
we therefore limit below our remarks to aspects of the results reported in proposition 5 having no immediate counterpart in the comments contained in remark 4 . in the cases
( i ) respectively ( i ) of proposition 5 the rules giving the period of the time evolution of the _ generic _ root @xmath349 see ( [ jtilaa ] ) respectively ( [ jtilba ] ) , are fairly straightforward and generally yield rather small periods , unless @xmath350 respectively @xmath351 with @xmath352 an _ irrational _ number _ positive _ but _ extremely small_. in cases ( ii ) and ( ii ) of proposition 5 the situation is quite interesting because the time evolution of the generic root @xmath128 can be either _ periodic _ with the basic period @xmath95 or
_ aperiodic_. note that in these cases the circle @xmath87 intersects the circle @xmath184 that is _
densely _ filled with _ square - root _
branch points , and moreover the branch point at @xmath194 ( which is now of _ irrational _ exponent , see ( [ branchzero ] ) ) is _ inside _
the circle @xmath353 this entails that , of the _ infinity _ of _ square - root _
branch points located on the piece of the circle @xmath184 that is _ inside _
@xmath320 either _ none _ , or _ all _ , are _
active_. the first case obtains if the root @xmath128 under consideration is _ initially _ on a sheet containing a branch point that does _ not _ fall inside @xmath320 hence the time evolution of this root @xmath130 $ ] as the point @xmath223 travels round and round on the circle @xmath87 brings it back to its point of departure after a single round ; equivalently , in this case the root @xmath354 $ ] belongs to a set of roots that does _ not _ get permuted as the point @xmath223 travels round and round on the circle @xmath87 . in the second case
the root @xmath128 under consideration starts from a sheet of the riemann surface that contains a branch point inside @xmath281 so that , when @xmath223 travels round and round on the circle @xmath87 , an _ endless _ sequence of _ different _ sheets get accessed by @xmath355 $ ] ; equivalently , such a root @xmath355 $ ] belongs to a set ( including an _ infinity _ of roots ) that does get permuted as the point @xmath223 travels round and round on the circle @xmath87 , with both mechanisms the pairwise exchange of some roots , and the cyclic permutation of an _ infinite _ number of roots playing a role at each round .
the identification of which sheets get thereby accessed , and in which order namely the specific shape of the trajectory when looked at , as it were stroboscopically , at the discrete sequence of instants @xmath356 @xmath357 is discussed in @xcite .
the extent to which this regime yields _ irregular _ motions is discussed further below , also to illuminate the distinction in these regimes between the time evolution entailed by our model with a given _ irrational _ value of @xmath140 and that of the analogous models with _ rational _ values of @xmath73 providing more and more accurate approximations of the given _ irrational _ value of @xmath73 . in case ( iii )
the time evolution is still _ isochronous _ , inasmuch as the results reported above entail that , for _ any _ given initial data ( excluding , of course , the _ special _ ones leading to a collision ; which are _ special _ in the same sense as a _ rational _ number is _ special _ in the context of _ real _ numbers ) , the motion of every root @xmath157 is _ periodic _ with one of the @xmath0 periods entailed by ( [ jtilc ] ) ( and note that the value of the integer @xmath358 provided by the _
third _ of these @xmath0 formulas is just the sum of the @xmath14 values for @xmath167 provided by the first @xmath14 of these @xmath0 formulas ) .
it is indeed clear that the initial data yielding such an outcome are included in an _ open _ set of such data , having of course _ full dimensionality _ in the space of initial data , _ all _ yielding the _ same _ outcome : since the periods do _ not _ change , see ( jtilc ) , if the change of the initial data , hence the change in the ratio @xmath359 is sufficiently tiny .
however the measures of these sets of data yielding the _ same _ outcome gets progressively _ smaller _ as the predicted periods get _ larger _ , and moreover the corresponding predictions involve more and more terms in the ( never ending ) _ continued fraction _ expansion of the irrational number @xmath339 , see ( [ confuna ] ) , displaying thereby , as @xmath323 increases towards unity , a _
progressively more sensitive _ dependence of the periodicity of our system on the initial data and moreover on the parameters ( the coupling constants , that determine the value of @xmath73 , see ( [ mu ] ) ) of our physical model ( [ eqmot]).@xmath240 0.3cm**example**. let us display here a specific example with the following ( conveniently chosen ) _ irrational _ value of @xmath73 in the interval @xmath360 ( hence corresponding to case ( iii ) of proposition 5 ) : @xmath361where of course @xmath362 is the _ golden ratio_. this assignment entails that _ all _ the coefficients @xmath363 , see ( [ confuna ] ) , are in this case unity , @xmath364 , hence the quantities @xmath365 see ( confune ) , coincide with the fibonacci numbers @xmath366,@xmath367and moreover one easily finds that @xmath368the corresponding formula for the periods obtains inserting these values in ( [ nukb ] ) and ( [ jtilc ] ) . from it , with some labor ,
one can obtain the following controlled estimate for the possible values of @xmath167:@xmath369these inequalities are valid for all values of @xmath323 ( in the interval @xmath370 ) ; they clearly entail that the integer @xmath167 diverges proportionally to @xmath371 as @xmath372 and that for @xmath373 with @xmath374the only possible values of @xmath167 are @xmath14 and @xmath375 this concludes our presentation of the results , and of some related observations , detailing the periodicity ( if any ) of the time evolution of a _ generic _ root @xmath128 of ( [ eqwtilde ] ) with ( [ ksi ] ) .
the identification of analogous , but of course more definite , results for the _ physical _ root @xmath376 and the consequential information on the periodicity ( if any ) of the solution of the physical problem ( eqmot ) as well as some additional information on the corresponding trajectories of the coordinates @xmath53 are provided in @xcite .
last but not least let us elaborate on the character of the _ aperiodic _ time evolution indicated under item ( ii ) of proposition 5 , including the extent it is _ irregular _ and it depends _ sensitively _ on its initial data .
it is illuminating to relate this question with the finding reported under item ( ii ) of proposition 4 , also in order to provide a better understanding of the relationship among the _ aperiodic _ time evolution that can emerge when @xmath73 is _ irrational _ ( see item ( ii ) of proposition 5 ) and the corresponding behavior say , with the same initial data for a sequence of models with _ rational _ values of @xmath73 ( see ( [ mupq ] ) ) that provide better and better approximations to that _ irrational _ value of @xmath73 ; keeping in mind the _ qualitative _ difference among the _ aperiodic _ time evolution emerging when @xmath73 is _ irrational _ , and the _ periodic _ indeed , even _ isochronous _ time evolutions prevailing whenever @xmath73 is _ rational _ , albeit with the qualifications indicated under item ( ii ) of remark 4 .
note that we are now discussing the case @xmath215 ( an analogous discussion in the @xmath151 case can be forsaken ) , with initial data such that the two circles @xmath184 and @xmath224 in the _ complex _
@xmath86-plane _ do intersect _ and moreover the origin @xmath194 falls _ inside _ the circle @xmath87 ( i. e. @xmath377 and @xmath378 let us then consider a given _ irrational _ value of @xmath215 and let the _ rational _ number @xmath379 ( with @xmath214 ) provide a _ very good _ approximation to @xmath140 which of course entails that the _ positive integers _ @xmath136 and @xmath64 are both _
very large_. consider then the _ difference _ @xmath380(see the second entry in ( [ jtilab ] ) ) between the _ two _ positive integers that characterize the _ two _ periods of the _ two _ time evolutions of @xmath128 corresponding to _ two _ sets of initial data that differ _ very little_. here clearly the quantity @xmath381 is the _ difference _ between the number of branch points that are enclosed inside the circle @xmath87 for these _ two _ different sets of initial data .
since the number @xmath64 of branch points on @xmath184is _ very large _ , it stands to reason that@xmath382where the _ quite small _ ( _ positive _ ) _ _ _ _ number @xmath383 provides a ( dimensionless ) measure of the difference between the _ two _ sets of initial data ( see for instance ( [ nu ] ) , which clearly becomes approximately applicable when @xmath64 is _ very large _ ) .
the _ floor _ symbol @xmath384 has been introduced in the right - hand side of this formula to account for the integer character of the numbers @xmath280 hence of their difference @xmath385 while the _ order of magnitude _ symbol @xmath386 indicates that the difference @xmath381 is proportional ( in fact equal , given the latitude left by our definition of the quantity @xmath387 to the quantity @xmath388 up to _ corrections _ which become _ negligibly small _ when @xmath383 is _ very small _ and @xmath64 is _ very large _ , but irrespective of the value of the quantity @xmath388 itself which , as the product of the _ large _ number @xmath64 by the _
small _ number @xmath383 , is required to be neither _ small _ nor _
large_. the relation@xmath389implied by this argument indicates that , for any given @xmath229 one can always choose ( finitely different ) initial data which differ by such a tiny amount that the corresponding periods are identical , confirming our previous statement about the _ isochronous _
character of our model whenever the parameter @xmath73 is _ rational_. but conversely this finding also implies that , for any set of initial data in the sector under present consideration ( i. e. that characterized by the inequalities @xmath390 and @xmath391 and by an additional specification to identify the physical root @xmath68 , see @xcite ) , if our physical model ( [ eqmot ] ) is characterized by an _
irrational _ value of @xmath140 see ( [ mu ] ) , and one replaces this value by a more and more accurate _ rational _ approximation of it , see ( mupq ) as it would for instance be inevitable in any numerical simulation corresponding to larger and larger values of @xmath136 and @xmath229 then one shall have to choose the two different assignments of initial data closer and closer to avoid a drastic change of period and for these very close sets of data the motion is indeed _ periodic _ with a period ( which we are able to predict , see ( [ jtilab ] ) and @xcite , but ) which becomes larger and larger the better one approximates the actual , _ irrational _ value of @xmath392 moreover in any numerical simulation the accuracy of the computation , in order to get the correct period , shall also have to increase more and more ( with no limit ) , because of the occurrence of closer and closer _ near misses _ through the time evolution ( associated with the coming into play of _ active _ branch points sitting on the circle @xmath184 closer and closer to the points of intersection with the circle @xmath87 ) . and
finally , if one insists in treating the problem with a truly _ irrational _
@xmath140 then , no matter how close the initial data are , the change in the periods becomes _ infinite _ because the difference @xmath393 in the number of _ active _ _ square - roots _ branch points on the circle @xmath184 included _ inside _ the circle @xmath87 is _ infinite _ ( see ( [ deltata ] ) ) , signifying that the motion is _ aperiodic _ , and that its evolution is indeed characterized by an _ infinite _ number of _ near misses _ , making it truly _ irregular .
_ this phenomenology , together with that of the _ near misses _ as described above , illustrates rather clearly the _ irregular _ character of the motions of our physical model when the coupling constants have appropriate values ( such as to produce an irrational value of @xmath73 outside the interval @xmath200 ) and the initial data are in the sector identified above .
note that the lyapunov coefficients associated with the corresponding trajectories vanish , because these coefficients as usually defined compare the difference ( after an _ infinitely long _ time ) of two trajectories that , to begin with , differ _ infinitesimally ; _ whereas our mechanism causing the _ irregular _ character of the motion requires , to come into play , an _ arbitrarily small but finite _ difference among the initial data .
the difference between these two notions corresponds to the fact that inside the interval between two _ _ differen__t real numbers
however close they may be there always is an _ infinity _ of _ rational _ numbers ; while this is _ not _ the case between two real numbers that differ only _ infinitesimally _ !
this observation suggests that , in an _ applicative _ context , the mechanism causing a _ sensitive dependence _ on the initial data manifested by our model may be _ phenomenologically _ relevant even when no lyapunov coefficient , defined in the usual manner , is _
positive_. as already observed previously @xcite , this mechanism is in some sense analogous to that yielding _ aperiodic _ trajectories in a triangular billiard with _ irrational
_ angles ; although in that case in contrast to ours this outcome is mainly attributable to the essentially singular character of the corners , and moreover no truly _ irregular _ motions emerge .
in this paper we have introduced and discussed a @xmath0-body problem in the plane suitable to illustrate a mechanism of transition from _ regular _ to _ irregular _ motions .
this model is the simplest one we managed to manufacture for this purpose .
its simplicity permitted us to discuss in considerable detail the mathematical structure underlining this phenomenology : this machinery can not however be too simple since it must capture ( at least some of ) the subtleties associated with the _ onset _ of an _ irregular _ behavior . therefore in this short paper we were only able to report our main findings without detailing their proofs , and we also omitted several other relevant aspects of our treatment ( including a fuller discussion of previous work by others in related areas ) : this material shall be presented in a separate , much longer , paper @xcite , and probably as well via an electronic version of our findings so as to supplement their presentation with various animations illustrating these results and their derivation . our main motivation to undertake this research project is the hunch that this mechanism of transition _ _ _ _ have a fairly general validity and be relevant in interesting applicative contexts .
hence we plan to pursue this study by focussing on other cases where this mechanism is known to play a key role , including examples ( see , for instance , @xcite and @xcite ) featuring a pattern of branch points covering densely an _ area _ of the _ complex _ plane of the independent variable rather than being confined just to reside densely on a _ line _ as is the case in the model treated herein ; and eventually to extend the application of this approach to problems of direct applicative interest . in this connection
the following final observation is perhaps relevant . in this paper as well as in others @xcite the main focus
has been on models featuring a transition from an _ isochronous _ to an _ irregular _ regime , and in this context much emphasis was put on the `` trick '' ( [ zita ] ) and in particular on the relationship it entails between the _ periodicity _ of the ( `` physical '' ) dependent variables @xmath93 as functions of the _ real _ independent variable @xmath6 ( `` time '' ) and the _ analyticity _ of other , related ( `` auxiliary '' ) dependent variables @xmath394 as functions of a _ complex _ independent variable @xmath31 .
but our findings can also be interpreted _ directly _ in terms of the _ analytic properties _ of the physical dependent variables @xmath53 as functions of the independent variable @xmath6 considered itself as a _ complex _ variable
. then the time evolution , which corresponded to a uniform travel round and round on the circle @xmath35 in the _ complex _ @xmath395-plane or equivalently on the circle @xmath87 in the _ complex _
@xmath86-plane , is represented as a uniform travel to the right along the _ real _ axis in the _ complex _
@xmath6-plane , while , via the relations ( see ( [ zita ] ) and ( [ ksi ] ) ) @xmath396the pattern of branch points in the _ complex _ @xmath31-plane or equivalently in the _ complex _ @xmath86-plane gets mapped into a somewhat analogous pattern in the _ complex _
@xmath6-plane , _ repeated periodically _ in the _ real _ direction with period @xmath34 see ( [ t ] ) .
in particular to mention the main features relevant to our treatment , see above the circle @xmath184 on which the _ square - root _ branch points in the _ complex _
@xmath86-plane sit , gets mapped in the _ complex _
@xmath6-plane into a curve @xmath397 on which sit the _ square - root _ branch points in the _ complex _ @xmath6-plane ; note that this curve @xmath397 ( in contrast to the circle @xmath184 ) _ does _ now _ depend _ on the initial data .
this curve is of course repeated periodically ; it is _ closed _ and contained in each vertical slab of width @xmath95 ( see figure [ fig2]b ) if the point @xmath88 is _ outside _ @xmath184 , otherwise it is _ open _ , starting in one slab and ending in the adjoining slab at a point shifted by the amount @xmath95 ; and it does not or does cross ( of course twice in each period ) the _ real _ axis in the _ complex _ @xmath6-plane depending whether , in the _ complex _
@xmath86-plane , the two circles @xmath184 and @xmath87 do not or do intersect each other ( see figure [ fig2]a ) . likewise ,
depending whether it is _ inside _ or _ outside _ the circle @xmath320 the point @xmath194 which , as entailed by our analysis , is a highly relevant branch point in the _ complex _
@xmath86-plane ( unless @xmath200 ) gets mapped into an analogous branch point located in each vertical slab _ above _ or _ below _ the _ real _ axis in the _ complex _ @xmath6-plane ; while the other branch point , at @xmath195 in the _ complex _
@xmath178-plane , gets mapped into an analogous branch point located at infinity in the _ lower half _ of the _ complex _ @xmath6-plane .
clearly the physical mechanism of _ near misses _ , which is the main cause of the eventual _ irregularity _ of the motion , becomes relevant only for initial data such that the curve @xmath397 crosses the real axis , thereby causing ( if @xmath73 is _ irrational _ ) an infinity of _ square - root _ branch points of the functions @xmath53 to occur _ arbitrarily close _ to the _ real _ axis in the _
_ complex__@xmath6-plane branch points which are however _ active _ ( namely , they actually cause a _ near miss _ in the physical evolution ) in only some ( yet still an infinity ) of the infinite number of vertical slabs in which the _ complex _ @xmath6-plane gets now naturally partitioned .
the _ near miss _ implies that the two particles involved in it slide past each other from one side or the other depending whether the corresponding branch point is just above or just below the real axis in the _ complex _ @xmath6-plane .
the _ sensitive _ dependence on the initial data is due to the fact that any tiny change of them causes some _ active _ branch point in the _ complex _
@xmath6-plane which is very close to real axis to cross over from one side of it to the other , thereby drastically changing the outcome of the corresponding _ near miss_. this terse discussion shows clearly that the explanation of the _ irregular _ behavior of a dynamical system in terms of travel on a riemann surface is by no means restricted to _
isochronous _ systems .
we found it convenient to illustrate in detail this paradigm by focussing in this paper on a simple _ isochronous _ model and by using firstly @xmath31 and then @xmath86 as independent _ complex _ variables but , as outlined just above , our analysis can also be done albeit less neatly by using directly the independent _ complex _ variable @xmath6 ; and the occurrence of a kind of periodic partition of the _ complex _
@xmath6-plane into an infinite sequence of vertical slabs
characteristic of our _ isochronous _ model does not play an essential role to explain the _ irregular _ character of the motion when such a phenomenology does indeed emerge .
the essential point is the possibility to reinterpret the time evolution as travel on a riemann surface , the structure of which is sufficiently complicated to cause an _
irregular _ motion featuring a _ sensitive dependence _ on its initial data .
the essential feature causing such an outcome is the presence of an _ infinity _ of branch points _ arbitrarily close _ to the _ real _ axis in the _ complex _
@xmath6-plane , the positions of which , as well as the identification of which of them are _ active _ , depends on the _ initial data _ nontrivially .
the model treated in this paper shows that such a structure can be complicated enough to cause an _ irregular _
motion , yet amenable to a simple mathematical description yielding a rather detailed understanding of this motion ; this suggests the efficacy also in more general contexts of this paradigm to understand ( certain ) _ irregular _ motions featuring a _ sensitive dependence _ on their initial data and possibly even to _ predict _ their behavior to the extent such a paradoxical achievement ( predicting the unpredictable ! ) can at all be feasible .
9 f. calogero , a class of integrable hamiltonian systems whose solutions are ( perhaps ) all completely periodic , _ j. math .
phys . _ * 38 * , 5711 - 5719 ( 1997 ) ; differential equations featuring many periodic solutions , in : l. mason and y. nutku ( eds ) , _ geometry and integrability _ , london mathematical society lecture notes , vol . * 295 * , cambridge university press , cambridge , 2003 , pp .
9 - 21 ; periodic solutions of a system of complex odes , _ phys .
lett . _ * a293 * , 146 - 150 ( 2002 ) ; on a modified version of a solvable ode due to painlev , _ j. phys .
gen . _ * 35 * , 985 - 992 ( 2002 ) ; on modified versions of some solvable odes due to chazy , _ j. phys .
gen . _ * 35 * , 4249 - 4256 ( 2002 ) ; solvable three - body problem and painlev conjectures , _ theor .
phys . _ * 133 * , 1443 - 1452 ( 2002 ) ; erratum * 134 * , 139 ( 2003 ) ; a complex deformation of the classical gravitational many - body problem that features a lot of completely periodic motions , _ j. phys . a : math . gen . _ * 35 * , 3619 - 3627 ( 2002 ) ; partially superintegrable ( indeed isochronous ) systems are not rare , in : _ new trends in integrability and partial solvability _ , edited by a. b. shabat , a. gonzalez - lopez , m. maas , l. martinez alonso and m. a. rodriguez , nato science series , ii .
mathematics , physics and chemistry , vol .
* 132 * , proceedings of the nato advanced research workshop held in cadiz , spain , 2 - 16 june 2002 , kluwer , 2004 , pp .
49 - 77 ; general solution of a three - body problem in the plane , _
j_. _ phys .
a : math . gen . _ * 36 * , 7291 - 7299 ( 2003 ) ; solution of the goldfish n - body problem in the plane with ( only ) nearest - neighbor coupling constants all equal to minus one half , _ _ j .
nonlinear math .
phys.__**11 * * , 1 - 11 ( 2004 ) ; two new classes of isochronous hamiltonian systems , _ j. nonlinear math .
phys_. * 11 * , 208 - 222 ( 2004 ) ; isochronous dynamical systems , _
applicable anal_. ( in press ) ; a technique to identify solvable dynamical systems , and a solvable generalization of the goldfish many - body problem , _ j. math . phys_. * 45 * , 2266 - 2279 ( 2004 ) ; a technique to identify solvable dynamical systems , and another solvable extension of the goldfish many - body problem , _ j. math .
phys_. * 45 * , 4661 - 4678 ( 2004 ) ; isochronous systems , _ proceedings _ of the conference on geometry , integrability and physics , varna , june 2004 ( in press ) ; isochronous systems , in : _ encyclopedia of mathematical physics _ , edited by j .-
p franoise , g. naber and tsou sheung tsun ( in press ) .
f. calogero and a. degasperis , novel solution of the integrable system describing the resonant interaction of three waves , _ physica d _ * 200 * , 242 - 256 ( 2005 ) .
f. calogero and j .- p .
franoise , periodic solutions of a many - rotator problem in the plane , _ inverse problems _ * 17 * , 1 - 8 ( 2001 ) ; periodic motions galore : how to modify nonlinear evolution equations so that they feature a lot of periodic solutions , _ j. nonlinear math .
phys . _ * 9 * , 99 - 125 ( 2002 ) ; nonlinear evolution odes featuring many periodic solutions , _ theor
phys . _ * 137 * , 1663 - 1675 ( 2003 ) ; isochronous motions galore : nonlinearly coupled oscillators with lots of isochronous solutions , in : _
superintegrability in classical and quantum systems _ , proceedings of the workshop on superintegrability in classical and quantum systems , centre de recherches mathmatiques ( crm ) , universit de montral , september 16 - 21 ( 2003 ) , crm proceedings & lecture notes , vol . *
37 * , american mathematical society , 2004 , pp . 15 - 27 ; new solvable many - body problems in the plane , _ annales henri poincar _ ( submitted to ) .
f. calogero , j .-
franoise and a. guillot , a further solvable three - body problem in the plane , _ j. math .
* 10 * , 157 - 214 ( 2003 ) .
f. calogero and v. i. inozemtsev , nonlinear harmonic oscillators , _ j. phys . a : math .
gen . _ * 35 * , 10365 - 10375 ( 2002 ) .
s. iona and f. calogero , integrable systems of quartic oscillators in ordinary ( three - dimensional ) space , _ j. phys . a : math .
gen . _ * 35 * , 3091 - 3098 ( 2002 ) . m. mariani and f. calogero , isochronous pdes , _ yadernaya fizika _
( russian journal of nuclear physics ) * 68 * , 958 - 968 ( 2005 ) ; a modified schwarzian korteweg de vries equation in 2 + 1 dimensions with lots of periodic solutions , _ yadernaya fizika_(in press ) .
m. sommacal , `` studio di problemi a molti corpi nel piano con tecniche numeriche ed analitiche '' , dissertaion for the `` laurea in fisica '' , universit di roma `` la sapienza '' , 26 september 2002 .
f. calogero and m. sommacal , periodic solutions of a system of complex odes .
higher periods , _ j. nonlinear math .
phys . _ * 9 * , 1 - 33 ( 2002 ) .
f. calogero , j .-
franoise and m. sommacal , periodic solutions of a many - rotator problem in the plane .
ii . analysis of various motions , _ j. nonlinear math .
* 10 * , 157 - 214 ( 2003 ) .
m. d. kruskal and p. a. clarkson , the painlev - kowalewski and poly - painlev tests for integrability , _ studies appl . math .
_ * 86 * , 87 - 165 ( 1992 ) .
r. d. costin and m. d. kruskal , nonintegrability criteria for a class of differential equations with two regular singular points , _ nonlinearity _ * 16 * , 1295 - 1317 ( 2003 ) .
r. d. costin , integrability properties of a generalized lam equation ; applications to the hnon - heiles system , methods appl
. anal . * 4 * , 113 - 123 ( 1997 ) .
t. bountis , l. drossos and i. c. percival , non - integrable systems with algebraic singularities in complex time , _
phys . a : math .
_ _ * * 23 * * , 3217 - 3236 ( 1991 ) .
t. bountis , investigating non - integrability and chaos in complex time , in : _ nato asi conf .
como , september 1993 & _ phys .
d _ * 86 * , 256 - 267 ( 1995 ) ; investigating non - integrability and chaos in complex time , _ physica _ * d86 * , 256 - 267 ( 1995 ) .
a. s. fokas and t. bountis , order and the ubiquitous occurrence of chaos , _ physica _ * a228 * , 236 - 244 ( 1996).s .
abenda , v. marinakis and t. bountis , on the connection between hyperelliptic separability and painlev integrability , j. phys .
a : math . gen . * 34 * , 3521 - 3539 ( 2001 ) .
e. induti , studio del moto nel piano complesso di @xmath398 particelle attirate verso l origine da una forza lineare ed interagenti a coppie con una forza proporzionale ad una potenza inversa dispari della loro mutua distanza " , dissertation for the laurea in fisica " , universit di roma la sapienza " , 26 may 2005 . | we introduce and discuss a simple hamiltonian dynamical system , interpretable as a @xmath0-body problem in the ( _ complex _ ) plane and providing the prototype of a mechanism explaining the transition from _ regular _ to _ irregular _ motions as travel on riemann surfaces .
the interest of this phenomenology illustrating the _ onset _ in a _
deterministic _ context of _ irregular _ motions is underlined by its generality , suggesting its eventual relevance to understand natural phenomena and experimental investigations . here
only some of our main findings are reported , without detailing their proofs : a more complete presentation will be published elsewhere . |
lep ii is very effective in bounding from below masses of charginos which , if this is kinematically allowed , should be produced in a pair in @xmath5 annihilation .
the present bounds are @xmath6 for the higgsino - dominated case and @xmath7 for the gaugino - dominated case if sneutrino is not too heavy @xcite . however ,
when the lightest chargino and neutralino ( the latter being the lsp ) are almost degenerate in mass , the charged decay products of the light chargino are very soft , and the above quoted bounds are no longer valid .
special search for such light charginos has been performed recently by delphi collaboration , and the case of @xmath8 is now excluded @xcite .
however , still in the case of @xmath9 lep ii does not provide a lower bound and charginos as light as @xmath10 are allowed ( this bound comes from the measurements of @xmath0 decays at lep i and slc ) .
the case of almost degenerate chargino and neutralino can be naturally realized in susy and the possibilities to find such particles are discussed in literature @xcite . in this letter
we investigate the radiative corrections to @xmath11 and to @xmath0 boson decay parameters generated by such almost degenerate particles .
when their masses are close to @xmath12 radiative corrections are large and they spoil the perfect description of experimental data by the standard model . due to the decoupling property of susy models ,
when @xmath13 the radiative corrections are power suppressed .
in the simplest supersymmetric extensions of the standard model the chargino - neutralino sector is defined by the numerical values of four parameters : @xmath14 , @xmath15 , @xmath16 and @xmath17 .
the case of nearly degenerate lightest chargino and neutralino naturally arise when : 1 .
[ case_a ] @xmath18 : in this case the particles of interest form an @xmath19 doublet of dirac fermions , whose wave functions are dominated by higgsinos ; 2 .
[ case_b ] @xmath20 : in this way we get an @xmath19 triplet of majorana fermions , with the wave functions dominated by winos . in this way
we get higgsino- and gaugino - dominated scenarios , correspondingly .
let us start from case ( [ case_a ] ) .
the degenerate dirac doublet produces the following corrections to the three functions @xmath21 which determine the values of radiative corrections ( for definitions of functions @xmath21 and further details about electroweak radiative corrections see @xcite ) : @xmath22 , \\
\delta^{\tilde h } v_a & = \frac{16}{9 } { \left } ( \frac{1}{2 } - s^2 + s^4 { \right } ) { \left } [ \frac{12 \chi^2 f(\chi ) - 2\chi - 1}{4\chi - 1 } { \right } ] , \\
\delta^{\tilde h } v_r & = \frac{16}{9 } c^2 s^2 { \left } [ { \left } ( 1 + 2\chi { \right } ) f(\chi ) - \frac{1}{3 } { \right}],\end{aligned}\ ] ] where @xmath23 , the function @xmath24 is defined in appendix b of ref .
@xcite , and @xmath25 ( @xmath26 ) is the sine ( cosine ) squared of the electroweak mixing angle @xmath27 . comparing the experimental data with standard model formulas @xcite , we obtain that the @xmath28 for the new physics contributions to @xmath21 can be computed in the following way : @xmath29 \label{eq40b } \begin{pmatrix } c_{mm } & c_{ma } & c_{mr } \\ c_{ma } & c_{aa } & c_{ar } \\
c_{mr } & c_{ar } & c_{rr } \end{pmatrix } = \begin{pmatrix } 7.28 & 0 & 0 \\ 0 & 7.24 & 2.54 \\ 0 & 2.54 & 23.03 \end{pmatrix } ; \qquad \begin{pmatrix } \overline{\delta v_m } \\
\overline{\delta v_a } \\
\overline{\delta v_r } \end{pmatrix } = \begin{pmatrix } -0.07 \\ -0.33 \\
+ 0.01 \end{pmatrix}.\end{gathered}\ ] ] in fig .
[ fig10 ] the functions @xmath30 are plotted against the chargino - neutralino mass @xmath31 . comparing this graph with formulas and
, we see that at @xmath32 c.l .
the bound @xmath33 should be satisfied .
note that the main contribution to @xmath28 comes from @xmath34 , which is singular at @xmath35 .
this singularity is not physical and our formulas are valid only for @xmath36 ; the existence of @xmath37 with a mass closer to @xmath12 will change z - boson breit - wigner curve , therefore it is also not allowed .
the importance of the @xmath0 wave function renormalization for the case of light charginos was emphasized in @xcite .
functions ( left y - axis ) and of the confidence level ( right y - axis ) on the light gaugino mass @xmath31 , in the limit @xmath18 ( higgsino - dominated case ) . ] for case ( [ case_b ] ) ( gaugino dominated states ) the expressions for the @xmath38 functions are : @xmath39 , \\ \delta^{\tilde w } v_a & = \frac{16}{9 } c^4 { \left } [ \frac{12 \chi^2 f(\chi ) - 2\chi - 1}{4\chi - 1 } { \right } ] ,
\\ \delta^{\tilde w } v_r & = \frac{16}{9 } c^2 s^2 { \left } [ { \left } ( 1 + 2\chi { \right } ) f(\chi ) - \frac{1}{3 } { \right}].\end{aligned}\ ] ] the values of these functions are shown in fig .
[ fig20 ] and at 95% c.l .
we get @xmath40 .
let us remark that , although this and the previous bounds have been obtained respectively in the limits @xmath41 and @xmath42 ( in which case the mass splitting @xmath43 is exactly zero ) , we have verified numerically using equations from ref . @xcite that they are still valid for values of @xmath44 and @xmath45 small enough to allow for @xmath9 .
functions ( left y - axis ) and of the confidence level ( right y - axis ) on the light gaugino mass @xmath31 , in the limit @xmath20 ( wino - dominated case ) . ]
since there are a number of new additional particles in susy extensions , we will briefly discuss their contributions to the functions @xmath21 . in the considered limits the remaining charginos and neutralinos are very heavy , so they simply decouple and produce negligible contributions .
the contributions of the three generations of sleptons ( with masses larger than @xmath46 ) into @xmath47 are smaller than @xmath48 , so they can be safely neglected .
the contributions of squarks of the first two generations are also negligible since they should be heavier than tevatron direct search bounds ; taking @xmath49 , we have @xmath50 .
concerning the contributions of the third generation squarks , they are enhanced by the large top - bottom mass difference and are not negligible . however , being positive and almost universal @xcite , they do not affect our analysis : compensating negative contributions of chargino - neutralino into @xmath47 they will generate positive contributions to @xmath51 and @xmath52 , and @xmath28 will not be better .
when squarks are heavy enough ( for @xmath53 ) , they simply decouple and their contributions become negligible as well .
the last sector of the theory to be discussed is higgs bosons . unlike the case of standard model
now we have one extra charged higgs and two extra neutral higgses .
their contributions to radiative corrections were studied in detail in @xcite . according to fig . 2 from that paper it is clear that the contributions of mssm higgses ( and @xmath54 gauge bosons ) equal with very good accuracy those of the standard model with the mass of sm higgs being equal to that of the lightest neutral higgs in susy generalization .
that is why the contributions from the gauge - higgs sector of the theory also can not compensate those of the light chargino - neutralino .
apart from oblique corrections ( those arising from vector bosons self energies ) which have been considered in this letter , there are process dependent vertex and box corrections .
however , due to lep ii and tevatron low bounds on squarks and sleptons masses they are small .
we are grateful to f. l. villante for useful discussion on @xmath28 analysis and to p. h. chankowski for bringing ref .
@xcite to our attention .
investigation of m. v. are supported by the grants of rfbr no .
98 - 07 - 90076 and 98 - 02 - 17372 . | though lep ii direct searches still can not exclude a chargino nearly degenerate with the lightest neutralino if their mass is only slightly above half of the @xmath0 boson mass , it can be excluded indirectly analyzing precision data . in this particular
limit simple analytical formulas for oblique electroweak radiative corrections are presented .
infn - fe-11 - 99 + hep - ph/9907219 @xmath1 dipartimento di fisica , universit di ferrara , i-44100 , ferrara , italy + @xmath2 istituto tesre / cnr , via gobetti 101 , i-40129 bologna , italy + @xmath3 infn , sezione di ferrara , i-44100 , ferrara , italy + @xmath4 itep , moscow , russia |
the experiments which are taking place at the cern lep ii electron positron collider will be able to explore the gauge structure of @xmath0 and @xmath1 three point functions in the @xmath4pair production @xcite . deviations from the standard model ( sm ) predictions for this reaction would indicate the existence of new physics effects .
in general , such deviations can be parametrized in terms of effective lagrangians by adding to the sm lagrangian , high dimension operators that describe the new phenomena @xcite . this model
independent approach accounts for new physics that shows up at an energy scale @xmath5 , larger than the electroweak scale .
the effective lagrangians are constructed with the light particle spectrum that exists at low energies , while the heavy degrees of freedom are integrated out .
they are invariant under the @xmath6 and , in the linearly realized version , they involve , in addition to the usual gauge boson fields , also the light higgs particle .
the most general dimension6 effective lagrangian , containing all sm bosonic fields , that is @xmath7 and @xmath8 even , was constructed in ref .
@xcite . in this letter
, we explore the consequence of new operators that give rise to anomalous higgs boson couplings , without affecting the self interaction amongst the gauge bosons .
these anomalous interactions can _ not _ be constraint by the lep ii results on w pair production .
in particular , we study the reactions @xmath9 and @xmath3 .
these processes do not involve any triple vector boson coupling , but only @xmath10 and @xmath11 anomalous higgs interactions . in the sm , neglecting the higgs electron coupling , there are no higgs contributions to these reactions at tree level since the @xmath10 and @xmath11 couplings are generated only at one
loop @xcite . out of the eleven independent operators constructed in ref.@xcite ,
two of them describe new interactions between the higgs particle and the vector bosons , @xmath12 where , in the unitary gauge , the higgs doublet becomes @xmath13^t$ ] , and @xmath14 , and @xmath15 , with @xmath16 and @xmath17 being the field strength tensors of the @xmath18 and @xmath19 gauge fields respectively .
we should notice that the operators ( [ lagrangian ] ) contribute only to the anomalous higgs couplings , @xmath20 , @xmath21 , since their possible contribution to the self
boson couplings , @xmath0 and @xmath1 , can be completely absorbed in the redefinition of the sm fields and gauge couplings @xcite .
furthermore , no new @xmath22 or @xmath3 vertices , like the ones studied in ref .
@xcite , are generated via these operators .
therefore , studies of anomalous trilinear gauge boson couplings are insensitive to the coefficients @xmath23 and @xmath24 .
anomalous higgs boson couplings have already been studied in higgs and z boson decays @xcite , in @xmath25 @xcite , @xmath26 @xcite , and @xmath27 colliders @xcite . here , in order to impose limits on the dimension6 operators ( [ lagrangian ] ) that generate the new higgs interactions , we examine the total @xmath22 and @xmath3 yield and kinematical distributions of the final state particles at lep ii and at the next linear collider , nlc .
in our analysis of the @xmath28 and @xmath29 reactions , we have taken into account the standard one loop higgs contributions @xcite to these processes . in this way
, the @xmath3 production involves seven ( eight ) standard ( anomalous ) feynman contributions , while in the @xmath22 process there are ten ( twelve ) standard ( anomalous ) diagrams . in order to compute these contributions ,
we have incorporated all anomalous couplings in helas
type @xcite fortran subroutines .
these new subroutines were used to adapt a madgraph @xcite output to include all the anomalous contributions .
we have checked that our code passed the non trivial test of electromagnetic gauge invariance .
we investigate the @xmath3 and @xmath22 production both at lep ii and at nlc , assuming a center of mass energy of @xmath30 and 500 gev and an integrated luminosity of @xmath31 and 50 fb@xmath32 respectively .
we required that the photon energy , @xmath33 , is larger than 5 gev for lep ii and than 20 gev for nlc .
we applied an isolation cut by requiring that the angle between any two final state particles is larger than @xmath34 .
the same cut was applied to the angle of the bosons with the beam pipe .
our results for the sensitivity of lep ii to the operators appearing in the effective lagrangian ( [ lagrangian ] ) indicate that the reaction @xmath35 can provide limits on both anomalous couplings since @xmath36 fb , and we can expect around 300 events per year .
a 1@xmath37 deviation in the total cross section , assuming a higgs boson with mass of 100 gev , requires that , @xmath38 these results are similar to those arising from the analysis of the reaction @xmath39 @xcite , where a 1@xmath37 deviation implies @xmath40 . on the other hand , the reaction @xmath41 is almost insensitive to the anomalous coefficient @xmath24 . for a higgs boson mass of @xmath42 gev
, a 1@xmath37 deviation can be observed for @xmath43 tev@xmath44 . the small sm cross section for this process , @xmath45 fb , leads to very few events per year at lep ii , and in consequence , the constraints from this reaction are limited by the poor statistics . due to the increase on the available phase space , we are able to establish tighter constraints on the coefficients @xmath46 by studying the contribution of the operators ( [ lagrangian ] ) to these same processes at nlc . for instance , at @xmath47 gev , @xmath48 fb , which yields 2900 events per year assuming the expected luminosity for nlc ( @xmath49 fb@xmath32 ) . requiring a maximum deviation of 2@xmath37 in the total cross section , and assuming a higgs boson of 200 gev , we obtain the allowed ranges @xmath50 tev@xmath44 and @xmath51 tev@xmath44 .
unlike at lep ii , at nlc the best constraint on the anomalous couplings come from the reaction @xmath52 , which gives rise to @xmath53 events per year at @xmath47 gev .
figure [ fig:1 ] shows the effect of the anomalous operators ( [ lagrangian ] ) in the total cross section of this reaction .
a 2@xmath37 deviation in the total cross section sets the bounds , @xmath54 also for a 200 gev higgs boson .
furthermore , we can expect the new interactions to affect most the longitudinally polarized gauge bosons production due to the presence of the scalar higgs boson as intermediate state .
we tried to take advantage of this fact by studying the longitudinal @xmath55pair production which brings on a slightly better result , _
@xmath56 tev@xmath44 and @xmath57 tev@xmath44 for a 2@xmath37 effect .
unfortunately the improvement is very small because the requirement of polarized @xmath55 reduces by two orders of magnitude the total yield .
we have also investigated different distributions of the final state particles in order to search for kinematical cuts that could improve the nlc sensitivity .
the most promising variable is the photon transverse momentum which distribution is presented in fig . [ fig:2 ] for the unpolarized case .
we observe that the contribution of the anomalous couplings is larger in the high @xmath58 region .
therefore a cut of @xmath59 gev drastically reduces the background .
the improvement on @xmath46 bounds can be clearly seen from fig .
[ fig:3 ] where a 1 , 2 , and 3@xmath37 deviations in the total cross section is shown before and after the above cut is applied .
the contribution of the anomalous couplings is dominated by on mass shell higgs production with the subsequent @xmath60 decay .
the peak around @xmath61 gev in the photon transverse momentum distribution is due to the monochromatic photon with @xmath62 which is emitted in association with the 200 gev higgs boson .
therefore the best constraints are obtained at nlc for higgs boson masses in the range @xmath63 gev , where a on shell production is allowed .
we present in table [ tab:1 ] the limits on the coefficients @xmath24 and @xmath23 based on a @xmath64 deviation in the total cross section for a higgs mass in the range @xmath65 gev .
table [ tab:2 ] shows the same limits after the cut in the photon transverse momentum distribution of @xmath59 gev is implemented .
these results are comparable to the limits for @xmath24 and @xmath23 obtained from the study of the reaction @xmath39 presented in ref .
@xcite , suggesting that the processes considered here can be used as a complementary tool in the study of anomalous higgs interactions . in conclusion , the search for the effect of higher dimensional operators that give rise to anomalous higgs boson couplings may provide important information on physics beyond the standard model and should be pursued in all possible reactions .
we have studied here the triple vector boson production , @xmath22 and @xmath3 , in @xmath25 colliders focusing on the operators that generate only anomalous higgs boson couplings and can not be tested in the @xmath4pair production .
we established the limits that can be imposed at lep ii and nlc through the analysis of the deviations on the total cross section .
furthermore , we used the photon transverse momentum spectrum and polarization of the @xmath55 s to improve these constraints to few tev@xmath44 at the nlc .
we would like to thank m. c. gonzalez
garcia for very useful discussions .
this work was supported by conselho nacional de desenvolvimento cientfico e tecnolgico ( cnpq ) , and by fundao de amparo pesquisa do estado de so paulo ( fapesp ) .
h. aihara _ et al .
_ , summary of the working subgroup on anomalous gauge boson interactions of the dpf long - range planning study , to be published in _ electroweak symmetry breaking and beyond the standard model _ , edited by t. barklow , s. dawson , h. haber and j. siegrist , hep - ph/9503425 ; + z. ajaltuoni _ et al .
_ , triple gauge boson couplings " , in _ proceedings of the cern workshop on lep ii physics _ , edited by g. altarelli _ et al .
_ , cern 96 - 01 , vol .
1(1996 ) p.525 , hep - ph/9601233 ; + t. barklow _ et al .
_ , summary of the snowmass subgroup on anomalous gauge boson couplings , to appear in the _ proceedings of the 1996 dpf / dpb summer study on new directions in high - energy physics _ , june 25 july 12 ( 1996 ) , snowmass , co , usa , hep - ph/9611454
. for a review , see h. georgi , _ weak interactions and modern particle theory _
( benjamim / cummings , menlo park , 1984 ) ; j. f. donoghue , e. golowich and b. r.holstein , _ dynamics of the standard model _ ( cambridge university press , 1992 ) .
k. hagiwara , s. ishihara , r. szalapski and d.zeppenfeld , phys .
* b283 * ( 1992 ) 353 ; _ idem _ , phys .
* d48 * ( 1993 ) 2182 ; k. hagiwara , t. hatsukano , s. ishihara and r. szalapski , nucl .
* b496 * ( 1997 ) 66 .
p. mry , m. perrottet , and f. m. renard , z.phys .
* c38 * ( 1988 ) 579 ; a. barroso and l. bento , phys.rev .
* d38 * ( 1988 ) 2742 ; f. boudjema _ et al .
_ , in _ proceedings of @xmath25 collisions at 500 gev : the physics potencial _ , edited by p. zerwas
( desy report no .
92 - 123a , b , hamburg , germany , 1992 ) , p. 717 ; u. baur and e. l. berger , phys .
* d47 * ( 1993 ) 4889 .
k. hagiwara and m. l. stong , z. phys .
* 62 * ( 1994 ) 99 ; b. grzadowski and j. wudka , phys .
lett . * b364 * ( 1995 ) 49 ; g. j. gounaris , f. m. renard and n. d.vlachos , nucl .
* b459 * ( 1996 ) 51 ; w. killian , m.krmer and p. m. zerwas , phys .
* b381 * ( 1996 ) 243 .
.range of the allowed values of the coefficients @xmath24 and @xmath23 , in tev@xmath44 , for a @xmath64 deviation in the total cross section of the process @xmath66 at nlc . [ cols="^,^,^",options="header " , ] | the effect of new operators that only modify the bosonic couplings of the higgs boson , without altering the @xmath0 or @xmath1 three point functions , are examined in the @xmath2 and @xmath3 processes .
we analyse the constraints on these interactions that can be imposed by the lep ii collider at cern and at the next linear collider . |
the gluon propagator is a fundamental green s function of quantum chromodynamics ( qcd ) .
certainly , a good description of this two point function is required to understand the non - perturbative regime of the theory . moreover , in the landau gauge , the gluon and the ghost propagators at zero momentum are connected with possible confinement mechanisms . in particular , the kugo - ojima confinement mechanism requires an infinite zero momentum ghost propagator @xcite and the zwanziger horizon condition @xcite implies either reflection positivity violation or a null zero momentum gluon propagator . from this last condition it can be proved that the zero momentum ghost propagator should diverge @xcite .
the computation of the gluon and ghost propagators for the full range of momenta can not be done in perturbation theory . presently , we have to rely either on dyson - schwinger equations ( dse ) or on the lattice formulation of qcd ; both are first principles approaches . in the first technique a truncation of an infinite tower of equations , together with the parametrizations of a number of vertices , namely the three gluon and the gluon - ghost vertices , is required to solve the equations . in the second , one has to care about finite volume and finite lattice spacing effects .
hopefully , the two solutions will be able to reproduce the same solution . in this article
we are concerned with the lattice landau gauge gluon propagator in momentum space , @xmath4 for pure gauge theory .
the form factor @xmath5 is the gluon dressing function .
recently , in @xcite the dse were solved for the gluon and ghost propagators in the deep infrared region .
the solution assumes ghost dominance and is a pure power law @xcite , with the exponents of the two propagators related by a single number @xmath1 . for the gluon dressing function
, the solution reads @xmath6 the dse equations allow a determination of the exponent @xmath7 , which , for the zero momentum , implies a null ( infinite ) gluon ( ghost ) propagator , in agreement with the criteria described above , and a finite strong coupling constant defined from the gluon and ghost dressing functions @xcite .
renormalization group analysis based on the flow equation @xcite were able to predict , for the solution ( [ zdse ] ) , the range of possible values for the exponent .
the results of this analysis being @xmath8 , implying , again , a null ( infinite ) zero momentum gluon ( ghost ) propagator .
a similar analysis of the dse but using time - independent stochastic quantisation @xcite predicted the same behaviour and @xmath9 . the reader should be aware that not all solutions of the dse predict a vanishing zero momentum gluon propagator . in @xcite , the authors computed a solution of the dse with a finite zero momentum gluon propagator .
moreover , in @xcite it was claimed that the gluon and ghost propagators in the deep infrared region are not connected via the same exponent @xmath1 and that there is nt yet a clear theoretical answer concerning the value of zero momentum gluon propagator . according to the authors , the range of possible values goes from zero to infinity . on the lattice , there are a number of studies concerned with the gluon propagator @xcite . on large lattices ,
the propagator was investigated in @xcite .
although the lattices used had a limited access to the deep infrared region , the authors conclude in favour of a finite zero momentum gluon propagator , which would imply @xmath10 for the solution ( [ zdse ] ) . in @xcite , the infinite volume and continuum limits of the landau gauge gluon propagator were investigated using various lattices and , again , the data supported a finite zero momentum gluon propagator . in @xcite , the three - dimensional lattice su(2 )
landau gauge propagator was studied with the authors measuring an infrared exponent compatible with the corresponding dse solution . in order to access the deep infrared region , in @xcite
we have computed the gluon propagator with large asymmetric lattices .
the fits to ( [ zdse ] ) produced always @xmath11 .
however , the inclusion of corrections to the leading behaviour given by ( [ zdse ] ) or when the gluon dressing function was modelled , either in the momenta region @xmath12 gev or for all the momenta , @xmath1 becomes larger than 0.5 .
moreover , in the preliminary finite volume study @xcite of the two point gluon function it came out that @xmath1 increases with the lattice volume and , in that sense , previously computed values should be regarded as lower bounds .
the previous studies do not provide a clear answer about the zero momentum gluon propagator . in this paper
, we update the results of our previous publications for the infrared region , using a larger set of lattices , that allows a better control of the infinite volume extrapolation . two different extrapolations to the infinite volume are performed , namely the extrapolation of @xmath1 and the extrapolation of the propagator , giving essentially the same result . the main conclusions from this investigation being that the lattice data is compatible with the dse solution ( [ zdse ] ) for the infrared gluon dressing function , for momenta @xmath13 mev or lower , and that the infinite volume extrapolation of the lattice gluon propagator seems to favour a vanishing zero momentum gluon propagator .
in the lattice formulation of qcd , the gluon fields @xmath14 are replaced by the links @xmath15 where @xmath16 are unit vectors along @xmath17 direction .
qcd is a gauge theory , therefore the fields related by gauge transformations @xmath18 are physically equivalent .
the set of links related by gauge transformations to @xmath19 is the gauge orbit of @xmath19 .
the gluon field associated to a gauge configuration is given by @xmath20 ~ - ~ \frac{1}{6 i g_0 } \mbox{tr } \big [ u_\mu ( x ) - u^\dagger_\mu ( x ) \big]\ ] ] up to corrections of order @xmath21 .
on the lattice , due to the periodic boundary conditions , the discrete momenta available are @xmath22 where @xmath23 is the lattice length over direction @xmath17 .
the momentum space link is @xmath24 and the momentum space gluon field @xmath25 \ , - \ , \frac{1}{3 } \mbox{tr } \big [ u_\mu ( \hat{q } ) - u^\dagger_\mu ( - \hat{q } ) \big ] ~ \bigg\ } ~ .\end{aligned}\ ] ] the gluon propagator is the gluon two point correlation function .
the dimensionless lattice two point function is @xmath26 on the continuum , the momentum space propagator in the landau gauge is given by @xmath27 assuming that the deviations from the continuum are negligable , the lattice scalar function @xmath28 can be computed directly from ( [ propcont ] ) as follows @xmath29 ~ \rangle \ , , \hspace{0.3 cm } q \ne 0 \ , , \label{dq2}\ ] ] and @xmath30 ~ \rangle \ , , \hspace{0.3 cm } q = 0 \ , , \label{dq20}\ ] ] where @xmath31 @xmath32 is the dimension of the group , @xmath33 the number of spacetime dimensions and @xmath34 is the lattice volume .
on the continuum , the landau gauge is defined by @xmath35 this condition defines the hyperplane of transverse configurations @xmath36 it is well known @xcite that @xmath37 includes more than one configuration from each gauge orbit . in order to try to solve the problem of the nonperturbative gauge fixing , gribov suggested the use of additional conditions , namely the restriction of physical configurational space to the region @xmath38 \ , \geq \ , 0 \ } ~ \subset ~ \gamma \ , , \ ] ] where @xmath39 ~ \equiv ~ - \nabla \cdot d[a ] $ ] is the faddeev - popov operator .
however , @xmath40 is not free of gribov copies and does not provide a proper definition of physical configurations .
a suitable definition of the physical configurational space is given by the fundamental modular region @xmath41 , the set of the absolute minima of the functional @xmath42 ~ = ~ \int d^{4}x ~ \sum_{\mu } ~ \mbox{tr}\left[a_{\mu}^{g}(x)a_{\mu}^{g}(x)\right ] \ , .
\label{fcont}\ ] ] in this article , the problem of gribov copies lattice , the gluon propagator was computed using both the gauge fixing method described here and the gauge fixing method described in @xcite , which aims to find the absolute maxima of @xmath43 $ ] ( see equation ( [ f ] ) ) . for a similar number of configurations ,
i.e. 164 configurations at @xmath44 , we found no clear differences in the propagators @xcite .
therefore , in this study it will be assumed that gribov copies do not play a significant role .
note that in @xcite , it was shown that , in the continuum , the expectation values measured in @xmath40 are free of gribov copies effects . ] on the computation of the gluon propagator will not be discussed . for a numerical study on gribov copies
see @xcite and references therein . on the lattice ,
the situation is similar to the continuum theory .
the landau gauge is defined by maximising the functional @xmath45 ~ = ~ c_{f}\sum_{x,\mu } \
, \mbox{re } \ , \ { \ , \mbox{tr } \ , [ g(x)u_{\mu}(x)g^{\dagger } ( x+\hat{\mu } ) ] \ , \ } \label{f}\ ] ] where @xmath46 is a normalization constant .
let @xmath47 be the configuration that maximises @xmath48 $ ] on a given gauge orbit .
for configurations near @xmath47 on its gauge orbit , we have @xmath49 = f_u [ 1 ] + \frac{c_f}{4 } \sum_{x,\mu } i \omega^a ( x ) \mbox{tr } \big [ & & \lambda^a \left ( u_\mu ( x ) \ , - \ , u_\mu ( x - \hat{\mu } ) \right ) ~ - \nonumber \\ & & \lambda^a \left ( u^\dagger_\mu ( x ) \ , - \ , u^\dagger_\mu ( x - \hat{\mu } ) \right ) \big ] \ , , \end{aligned}\ ] ] where @xmath50 are the gell - mann matrices . by definition
, @xmath47 is a stationary point of @xmath51 , therefore @xmath52 ~ = ~ 0 \ , .
\label{statf}\end{aligned}\ ] ] in terms of the gluon field , this condition reads @xmath53 \ , + \ , \mathcal{o } ( a^2 ) ~ = ~ 0 \ , , \ ] ] or @xmath54 i.e. ( [ statf ] ) is the lattice equivalent of the continuum landau gauge condition .
the lattice faddeev - popov operator @xmath55 is given by the second derivative of ( [ f ] ) .
similarly to the continuum theory , on the lattice one defines the region of stationary points of ( [ f ] ) @xmath56 the gribov s region @xmath40 of the maxima of ( [ f ] ) , @xmath57 and the fundamental modular region @xmath58 defined as the set of the absolute maxima of ( [ f ] ) .
a proper definition of the lattice landau gauge chooses from each gauge orbit , the configuration belonging to the interior of @xmath58 . in this work ,
the gauge fixing algorithm used is a fourier accelerated steepest descent method ( sd ) as defined in @xcite . in each iteration ,
the algorithm chooses @xmath59 \ , - \ , \mbox{trace } \right ) \bigg ] \label{sd}\ ] ] where @xmath60 @xmath61 are the eigenvalues of @xmath62 , @xmath63 is the lattice spacing and @xmath64 represents a fast fourier transform ( fft ) . for numerical purposes ,
it is enough to expand to first order the exponential in ( [ sd ] ) , followed by a reunitarization of @xmath65 . on the gauge fixing process ,
the quality of the gauge fixing is measured by @xmath66 \label{theta}\ ] ] where @xmath67\ ] ] is the lattice version of @xmath68 .
.lattice setup used in the study of the gluon propagator volume dependence . [ cols=">,>,>,>,>",options="header " , ] in the following , we will not consider extrapolations using all lattices ( set i ) , because it involves our smallest lattice , which is too short @xmath69 fm in the spatial directions . note , however , that the data reported in table [ chivinfty ] point towards a smooth approach to the infinite volume limit , starting from lattices sizes as small as 0.8 fm .
furthermore , figures [ d01v ] and [ d23v ] suggest that we should not use a linear extrapolation . according to the figures ,
a quadratic extrapolation , or a higher power of the inverse volume , seems to be more suitable .
however , for set ii , and for the first non - vanishing momentum , the cubic extrapolation gives a quite poor description of the data , when compared to the quadratic fit .
so , for the reasons explained above , from now on we will consider only the quadratic extrapolation using set ii and iii . figures [ d01v ] and [ d23v ] show the lattice data together with the two quadratic extrapolations to the infinite volume .
note that in both extrapolations , the worst @xmath70 is associated with the first nonzero momentum .
indeed , as can be seen in figure [ d01v ] , for this momentum the data seems to fluctuate more than for the other momenta . for larger momenta the behaviour of the gluon propagator as a function of
the volume becomes smoother .
in @xcite the infinite volume of the renormalized zero momentum gluon propagator was computed .
the chosen renormalization condition was @xmath71 the lattice data was renormalized at @xmath72 gev .
the quoted extrapolated renormalized zero momentum gluon propagator being @xmath73 gev@xmath74 .
our largest momentum available is @xmath75 gev , a slightly lower value . performing the renormalization in the same way but at the scale @xmath76 gev
, it comes @xmath77 the quoted errors are pure statistical .
the value quoted in @xcite is in between the above figures , with the result from the so called set iii being almost compatible within one standard deviation with @xmath73 gev@xmath74 and the result from set ii being compatible only at three standard deviations .
the extrapolated propagators can be seen in figure [ dext ] ; the errors were computed assuming gaussian error propagation . for infrared momenta , the extrapolated propagators interpolate between the two lattices simulated in @xcite . moreover , for zero momentum the extrapolated propagators are smaller than the propagator of leinweber _ et al _ computed with their largest lattice .
the extrapolated propagators from fitting the two different sets of lattices are compatible , at least , at the level of two standard deviations .
the propagator computed from set iii shows much larger statistical errors .
the differences between the two propagators are larger for smaller momenta .
this difference can be used to estimate systematic errors .
the fit of the dressing function computed from the extrapolated propagator , obtained from set ii , to the first three nonzero momenta to the pure power law gives @xmath78 and has a @xmath79 .
similar fits but using the corrections to ( [ zdse ] ) produce larger values for the same exponent ( @xmath80 ) .
all the computed @xmath1 s are compatible with a vanishing zero momentum gluon propagator .
moreover , the fit to the pure power law produces a number which agrees well with the estimation from an extrapolation on @xmath1 - see equations ( [ kinfty ] ) and ( [ kinfty2 ] ) . if , instead , we perform the same analysis but for the extrapolated propagator obtained from set iii , then @xmath81 and @xmath82 .
again , the fits with corrections to the pure power law give large values for @xmath83 .
note that now the @xmath1 fitting the pure power law is smaller than the corresponding value obtained from our largest lattice ( see table [ kappa ] ) and is compatible with @xmath78 within two and a half standard deviations .
and @xmath84 propagators computed at @xmath44 in @xcite . ] and @xmath84 propagators computed at @xmath44 in @xcite . ]
we have computed the landau gauge gluon propagator for large asymmetric lattices with different volumes . for each volume
, we have checked that the lattice dressing function is well described by the dse solution ( [ zdse ] ) for momenta below @xmath13 mev by fitting the above functional form . the exception being our smallest lattice @xmath85 .
note that in all these fits @xmath86 was never used . in what concerns the infrared exponent @xmath1
, the data shows finite volume effects , with @xmath1 becoming larger for larger volumes . in this sense , all values in table [ kappa ] can be read as lower bounds on the infinite volume figure .
the infinite volume @xmath1 was estimated in two different ways : i ) from an extrapolation of the values reported in table [ kappa ] , assuming a linear and a quadratic dependence on the inverse volume ; ii ) extrapolating the propagator , assuming a quadratic dependence on @xmath87 .
the results of the first method are ( [ kinfty ] ) , ( [ kinfty2 ] ) , giving an weighted mean value of @xmath88 .
the value for the exponent from extrapolating directly the gluon propagator being @xmath78 , if one considers the five largest lattices ( set ii ) , and @xmath81 , if one considers the four largest lattices ( set iii ) .
the first value is on the top of the value obtained from extrapolating directly @xmath1 as function of the volume .
moreover , these two values agree well with the time - independent stochastic quantisation prediction @xmath9 @xcite and are within the range of possible values estimated with renormalization group arguments @xcite , @xmath8 . all these results point towards a vanishing zero momentum gluon propagator .
the @xmath81 , obtained from fitting the gluon dressing function , computed from the quadratic extrapolation of the gluon propagator to the infinite volume , using only the four largest lattices ( set iii ) , is compatible with both an infinite and a null zero momentum gluon propagator .
this value is smaller than the measured exponent from our largest lattice , @xmath89 , and , although , these two numbers being compatible within errors , the extrapolated @xmath1 does not follow the observed behaviour that @xmath1 increases with the lattice volume .
this is probably due to the large statistical errors observed in the extrapolation using the smaller set of lattices .
finally , one can claim a @xmath2 with the lattice data favouring the right hand side of the interval .
although the lattice data seems to favour a null zero momentum gluon propagator , the measured extrapolated zero momentum propagator clearly does not vanish - see equation ( [ dr0 ] ) . as a function of the lattice volume ,
the bare zero momentum propagator becomes smaller for larger volumes .
moreover , in @xcite it was shown that the extrapolation of the zero momentum propagator , depending on the functional form considered , is compatible with both a vanishing and a nonvanishing value . in this study
we avoided the question of the zero momentum value by fitting only the smallest nonzero momenta .
solutions to this puzzle require simulations on larger lattices and/or having a better theoretical control on the volume extrapolation .
currently , we are engaged in improving the statistics for our larger lattices and simulating bigger volumes , aiming to improve the infinite volume extrapolations .
we would like to thank c. s. fischer , j. i. skullerud , a. g. williams , p. bowman , d. leinweber and a. c. aguilar for inspiring discussions .
j. silva acknowledges financial support from fct via grant sfrh / bd/10740/2002 .
99 t. kugo , i. ojima , _ prog .
* 66 * ( 1979 ) 1 .
t. kugo , hep - th/9511033 .
d. zwanziger , _ nucl .
phys . _ * b364 * ( 1991 ) 127 .
d. zwanziger , _ nucl .
* b378 * ( 1992 ) 525 .
d. zwanziger , _ nucl .
_ * b399 * ( 1993 ) 477 .
d. zwanziger , _ nucl .
phys . _ * b412 * ( 1994 ) 657 . c. lerche , l. von smekal , _ phys .
_ * d65 * ( 2002 ) 125006 [ hep - ph/0202194 ] . for reviews
see @xcite and @xcite .
r. alkofer , l. von smekal , _ phys .
rep . _ * 353 * ( 2001 ) 281[hep - ph/0007355 ] . c. s. fischer , hep - ph/0304233 . c. s. fischer , r. alkofer , _ phys .
_ _ b536 _ ( 2002 ) 177 [ hep - ph/0202202 ] . c. s. fischer , r. alkofer , h. reinhardt , _ phys .
* d65 * ( 2002 ) 094008[hep - ph/0202195 ] . j. m. pawlowski , d. f. litim , s. nedelko , l. von smekal , _ phys .
* 93 * ( 2004 ) 152002 [ hep - th/0312324 ] .
d. f. litim , j. m. pawlowski , s. nedelko , l. von smekal , hep - th/0410241 . c. s. fischer , h. gies , _ jhep _ 0410 ( 2004 ) 048 [ hep - ph/0408089 ] .
d. zwanziger , _ phys .
rev . _ * d65 * ( 2002 ) 094039 [ hep - th/0109224 ] .
d. zwanziger , _ phys .
* d67 * ( 2003 ) 105001 [ hep - th/0206053 ] .
a. c. aguilar , a. a. natale , p. s. rodrigues da silva , _ phys .
* 90 * ( 2003 ) 152001 [ hep - ph/0212105 ] .
a. c. aguilar , a. a. natale , _ jhep _ * 0408 * ( 2004 ) 057 [ hep - ph/0408254 ] .
boucaud , j. p. leroy , a. le yaouanc , a. y. lokhov , j. micheli , o , pne , j. rodrguez - quintero , c. roiesnel , hep - lat/0507005 .
boucaud , j. p. leroy , a. le yaouanc , a. y. lokhov , j. micheli , o , pne , j. rodrguez - quintero , c. roiesnel , hep - ph/0507104 .
j. e. mandula , m. ogilvie , _ phys .
* b185 * ( 1987 ) 274 .
j. e. mandula , _ phys .
* 315 * ( 1999 ) 273 [ hep - lat/9907020 ] .
a. cucchieri , t. mendes , a. taurines,_phys .
_ * d67 * ( 2003 ) 091502 [ hep - lat/0302022 ] .
j. c. r. bloch , a. cucchieri , k. langfeld , t. mendes , _ nucl .
_ * b687 * ( 2004 ) 76 [ hep - lat/0312036 ] .
s. furui , h. nakajima , _ phys .
* d69 * ( 2004 ) 074505 [ hep - lat/0305010 ] .
a. sternbeck , e .-
ilgenfritz , m. muller - preussker , a. schiller , _ phys .
_ * d72 * ( 2005 ) 014507 [ hep - lat/0506007 ] .
d. b. leinweber , j. i. skullerud , a. g. williams , c. parrinello , _ phys . rev .
_ * d60 * ( 1999 ) 094507 ; _ erratum _ , _ phys . rev . _
* d61 * ( 2000 ) 079901 [ hep - lat/9811027 ] .
f. d. r. bonnet , p. o. bowman , d. b. leinweber , a. g. williams , j. m. zanotti , _ phys .
_ _ d64 _ ( 2001 ) 034501 [ hep - lat/0101013 ] .
o. oliveira , p. j. silva , _ aip conf .
* 756 * ( 2005 ) 290 [ hep - lat/0410048 ] .
p. j. silva , o. oliveira , _ pos _
( lat2005 ) 286 [ hep - lat/0509034 ] .
o. oliveira , p. j. silva , _ pos _
( lat2005 ) 287 [ hep - lat/0509037 ] . | the infrared limit of the lattice landau gauge gluon propagator is studied .
we show that the lattice data is compatible with the pure power law @xmath0 solution of the dyson - schwinger equations . using various lattice volumes , the infinite volume limit for the exponent @xmath1
is measured .
although , the results allow @xmath2 , the lattice data favours @xmath3 , which would imply a vanishing zero momentum gluon propagator . |
as reported previously by a number of researchers , the brightness field of solar granulation bears evidence for the presence of features with lifetimes far exceeding those of granules .
in particular , roudier _ et al .
_ @xcite detected dark `` intergranular holes '' singularities in the network of intergranular lanes , which could be observed for more than 45 minutes and whose diameters range from 180 to 330 km .
similar features , which persisted over up to 2.5 hours , were also observed by hoekzema , brandt , and rutten @xcite and hoekzema and brandt @xcite .
a different form of long - term persistence , closely related to spatial structuring , was revealed in granulation images averaged over intervals of one to two hours . as noted by getling and brandt @xcite and getling ( 2006 , paper i ) , averaged brightness fields consist of bright , nearly granular - sized blotches against a darker background ; they may form long - lived , quasi - regular systems of concentric rings or parallel strips `` ridges '' and `` trenches '' in the relief of the brightness field . later , algorithmic treatment showed that patterns of alternating bright and dark lanes ( varying in the degree of their regularity ) are virtually ubiquitous in averaged granulation images ( getling and buchnev , @xcite ) . since the material upflows in the visible layers of the solar atmosphere are brighter than the downflows , such patterns should reflect the presence ( and , possibly , widespread occurrence ) of roll convection in the upper convection zone . in particular , roll systems could form the fine structure of larger - scale , supergranular and/or mesogranular , convective flows . some other indications of a long - term spatial organisation in the granulation field have also been revealed .
et al._@xcite found that granules with longer lifetimes exhibit a tendency to form mesogranular - scaled clusters .
muller , roudier , and vigneau @xcite also detected such clustering in the spatial arrangement of large granules ; they emphasised a plausible relationship between the clusters and mesogranules .
et al._@xcite reported their observations of a specific collective behaviour of families ( `` trees '' ) of fragmenting granules .
such families can persist for up to eight hours and appear to be related to mesogranular flows .
an imprint of the supergranular structure can also be traced in the granulation pattern ( baudin , molowny - horas , and koutchmy , @xcite ) . as a particular argument for the presence of long - lived features , we noted ( getling and brandt , @xcite ) a conspicuously slow decrease in the rms contrast of the averaged images with the averaging time .
in his comment on our original paper ( getling and brandt , @xcite ) , rast @xcite suggested that the quasi - regular structures detected in the granulation patterns are merely of statistical nature and do not reflect the structure of real flows .
he artificially constructed a series of random fields with some characteristic parameters typical of the solar granulation ( claiming nevertheless that this series `` is not an attempt to model solar granulation '' and represents `` a completely random and changing flow pattern '' ) and found that the law of contrast variation with the averaging time does not change substantially if the real images are replaced by such fields . on this basis , rast @xcite raised doubts on the real presence of a long - term component in the granulation dynamics . in view of attributing the features of regularity in the averaged images to statistical effects
, rast @xcite also put forward some other arguments . getling ( paper i ) discussed them in detail and has shown that the reasoning by rast @xcite does not contest the possibility of a physical origin of the quasi - regular structures .
at the same time , the particular issue of the contrast variation with the averaging time was reserved for a special consideration .
this paper continues the publication of our study whose first part was presented in paper i and addresses precisely the contrast - variation laws for various series of images . generally , the less variable the pattern , the slower the contrast decline with the length of the averaged sequence of images .
thus , the direct implication of a slow decrease in the contrast of averaged images is the presence of persistent features in the brightness field under study .
the contrast - variation law itself does not reflect the degree of spatial order in this field ; however , if the regularity is associated with temporal persistence , the presence of regular structures will flatten the contrast - variation curve . for these reasons
, we extend here the conceivable range of interpretation of our results compared to that discussed in paper i. accordingly , we have slightly modified the main title of the paper ( `` do quasi - regular structures really exist in the solar photosphere ? '' ) , which was originally intended to be common for both parts .
we compare here the laws of contrast variation with the averaging time for real granulation images , images obtained by numerical simulation of convection on a granular scale , the artificial fields constructed by rast @xcite , and a series of `` shuffled '' images ( obtained as subfields of the original series of granulation images taken at random positions in both spatial coordinates and at random times ) .
our results confirm the presence of a long - lived component in the granulation field , while the counterargument by rast @xcite based on the contrast variation can be dismissed .
by the real granulation images , we mean the images of the la palma series obtained by brandt , scharmer , and simon ( see simon _ et al . _ , @xcite ) on 5 june 1993 using the swedish vacuum solar telescope ( la palma , canary islands ) .
we analysed a seven - hour subset of this series ; the frame cadence was 21.03 s , and a @xmath5 mm@xmath6 subfield ( @xmath7 pixels with a pixel size of 0.125@xmath8 , or 90.6 km ) was cut out of the original images . in paper
i , we already described some details of the technique of data acquisition and pre - processing employed by simon _ et al . _
@xcite .
the random fields constructed by rast @xcite are additive superpositions of @xmath9 randomly disposed two - dimensional gaussians that vary randomly in amplitude and radius around given mean values .
these values and the time scale of the evolution were chosen so as to mimic the corresponding parameters of the solar granulation .
the sequence of such fields represents a continuous time evolution with the persistent emergence of `` new '' gaussian peaks and disappearance of `` old '' ones .
the resulting images contain @xmath10 pixels . in the language of the numerical correspondence between the parameters of the random field and those of the granulation patterns , the images follow at a time step of 20
s over a period of 8.25 hours .
another sequence of random fields was prepared from a series of real ( speckle - reconstructed ) granulation images obtained with the dutch open telescope ( dot ) , la palma , on 19 october 2001 .
this series comprised 198 frames of @xmath11% rms contrast , measuring @xmath12 pixels with a pixel size of 0.071 arcsec and taken in the 432 nm continuum with a time lag of 30 s between frames . to generate randomised fields ,
the images were subject to the following shuffling procedure .
three series of random numbers ranging between 0.0 and 1.0 were generated by an appropriate standard computer program .
the numbers in two of these series were used to extract randomly - shifted subfields of @xmath13 pixels from the original field .
the shifts were taken as the above - mentioned random numbers multiplied by the maximum shifts , 600 pixels in @xmath14 and 400 in @xmath15 .
the third random number served to compose a new sequence of 198 frames by randomly shuffling the sequence number of the selected frames , from which the random subfield was extracted .
in addition to the la palma granulation images , the random fields constructed by rast @xcite , and the sequence of shuffled granulation images , we consider here two series of artificial granulation patterns obtained by numerical simulation of granular - scaled convection .
the first one was prepared by rieutord _
et al . _
an image of this set is made of @xmath16 pixels ( a single pixel being 95.24 km ) and covers a @xmath17 mm@xmath6 area .
the interval between images corresponds to 20 s of real time .
the second series was computed by m. steffen , b. freytag , and h .-
ludwig ( see wedemeyer _
_ , @xcite ) for an @xmath18 mm@xmath6 area ; each image is represented by @xmath19 pixels with a pixel size of 56 km and the interval between frames is 30 s.
we note that the rms contrast of the individual images in the la palma series varies widely , and even the averaged images exhibit a large spread in the contrast values , depending on the central time of the averaging interval .
this raises considerable difficulties as the laws of contrast variation with the averaging time are compared between the real granulation , the artificial field ( rast @xcite ) , and simulated granulation ( as noted in the introduction , it is on such comparisons that rast @xcite largely based his criticism on the inferences of the original paper by getling and brandt @xcite ) . for this reason , to obtain unambiguous results ,
we normalize both the mean intensity and the rms intensity contrast of each image to a standard value for either of these two quantities .
we apply this procedure to the four above - mentioned data series ( the normalisation of the dot series was part of the speckling technique ) .
in addition , we remove the residual mean - intensity gradient within each frame ( such gradients were noticeable in the images of the la palma series ) . for the data series
so normalised , we compute the variation in the rms contrast with the averaging time @xmath1 ( at a fixed mid - time of the averaging interval and with a maximum averaging time of about 7 hours ) . in the case of the shuffled images ,
we calculate the average of @xmath0 images , from the first to the @xmath0-th one , the contrast of this average , and the coefficient of correlation between the first and the @xmath0-th image ; we represent both the contrast and the correlation coefficient as a function of @xmath0 .
the resulting contrast - variation curves are plotted in figure [ contr ] , on both linear ( figure [ contr]a ) and log log ( figure [ contr]b ) scales .
the rms contrast of an individual image is set to unity in all cases ; in this sense , the contrast values presented here can be called normalised .
for comparison , in figure [ contr]b , we also show a straight line that corresponds to the statistical @xmath3 law of decrease in the rms contrast of @xmath0 averaged completely random fields with @xmath0 , or , in other words , to the @xmath2 law of contrast decrease with the averaging time @xmath1 .
it can be seen that the curve for the real solar images runs well above the other three curves .
it diverges most widely with the curve for the random field constructed by rast @xcite , while the curves for the numerically simulated granulation remain in between these extremes .
although the differences in the slope are most pronounced in the region of @xmath20 h , they remain quite appreciable over the entire range of averaging times .
most instructive are comparisons of the slopes of the contrast - variation curves with the slope of the @xmath2 straight line in figure [ contr]b .
the contrast of the real granulation images decreases with @xmath1 more slowly than according to the @xmath2 law even in the rightmost part of the graph , at averaging times of several hours . at the same time , the curve for the field constructed by rast @xcite is nearly parallel to the @xmath2 line for @xmath21 hours . a similarly rapid decline
can also be noted for one of the series of simulated images ( wedemeyer _ et al . _
@xcite ) ; it is virtually consistent with the statistical law in the region of @xmath22 hour .
as for the images obtained by rieutord _
et al . _
@xcite in their simulations , the corresponding curve also runs well below that of the la palma series , although , among the four curves in figure [ contr ] , it exhibits the slowest decline at @xmath23 hours , even approaching the la palma curve at large @xmath1 .
the fact that the relatively slow decline in the contrast values should obviously be associated with the presence of some long - lived component in the granulation field is additionally illustrated by figure [ contr_shuffl ] , which shows the variation in the rms contrast of averaged sequences of @xmath0 shuffled granulation images as a function of the number @xmath0 .
this variation is plotted on a linear scale in figure [ contr_shuffl]a and on a log log scale in figure [ contr_shuffl]b .
in addition , figure [ contr_shuffl]a presents the coefficient of correlation between the first and the @xmath0th frame in the sequence as a function of @xmath0 .
a straight line corresponding to the statistical @xmath3 law is also plotted for comparison in figure [ contr_shuffl]b .
two remarkable features can be noted immediately .
first , the correlation coefficient of the shuffled images is very small even at very small @xmath0 ( _ e.g. _ , @xmath4 ) and does not exhibit any systematic trend , fluctuating about zero throughout the entire interval of @xmath0 .
this means that the shuffled images can be considered to be completely uncorrelated .
second , in agreement with the first property , they follow very closely the @xmath3 over the whole interval of @xmath0 , again starting from small @xmath0 , such as @xmath4 ( the very small depression of the curve in the range @xmath24 can be attributed to the incomplete validity of the statistical law at such moderate values of @xmath0 ) .
it is noteworthy that the field constructed by rast @xcite , as noted above , follows the @xmath3 law only from averaging times of about 0.3 hours , while it exhibits a slower contrast decline at shorter averaging times ; obviously , this effect results from the finite lifetimes of the brightness peaks in the field constructed by rast @xcite , which makes the images within these lifetimes correlated rather than statistically independent .
in contrast , the shuffling procedure appears to completely destroy any long - lived features that might be present in the original series .
we see that the contrast of the averaged images of solar granulation decreases with the averaging time considerably more slowly than does the contrast of the averages of either a random field similar to the granulation in some of its parameters ( rast , @xcite ) , or sequences of shuffled granulation images , or the granulation pattern numerically simulated by steffen and his colleagues ( see wedemeyer _
et al._@xcite ) .
this difference is especially pronounced at short to moderate averaging times . the artificial granulation pattern obtained in simulations by rieutord _ et al .
_ @xcite differs markedly from the other computed pattern ( wedemeyer _ et al . _
@xcite ) in the behaviour of the contrast of averaged images .
the former series of images exhibits , at @xmath25 hours , the slowest decline in the rms contrast with the averaging time @xmath1 ( among all the five datasets considered here ) and thus appears to contain a persistent component , while the latter demonstrates a contrast variation consistent with the statistical @xmath2 law at sufficiently large @xmath1 .
this fact remains currently unexplained . in principle , effects of the sort could be attributed to an implicit , uncontrolled influence of the boundary conditions used in the simulations ; these conditions can favour the persistence of some features in the simulated velocity field .
however , checking this conjecture would require a highly scrupulous analysis of the numerical model . in the context of the adequacy of the so - called realistic simulations of solar convection
, it would be worth doing such an analysis in the future .
the results of our comparative analysis of the contrast - variation laws indicate quite definitely that the brightness field of solar granulation contains a long - lived component .
our approach can not distinguish between particular forms of this component .
one of us ( p.n.b ) adheres to the point of view that it is formed by locally persistent dark intergranular holes ( roudier _ et al .
_ , @xcite ; hoekzema _ et al .
_ , @xcite ; hoekzema and brandt , @xcite ) . the other author of this paper ( a.v.g . )
is inclined to attribute the observed contrast - variation law to the presence of quasi - regular structures ( first reported by getling and brandt @xcite and discussed in detail by getling in paper i ) .
most likely , it should not be assumed that these two interpretations contradict each other , since features similar to intergranular holes can naturally be present in quasi - regular structures .
our comparison between the contrast - variation curves for the real granulation and the random field constructed by rast @xcite completes the discussion in paper i concerning the critical comments by rast @xcite on our original paper ( getling and brandt , @xcite ) .
based on all our reasoning , we can dismiss the suggestion by rast @xcite on the purely statistical nature of the quasi - regular structures in the granulation field .
we are obliged to r.a .
shine and p. stterlin for making the soho mdi and dot data , respectively , available to us .
we thank t. roudier , m. steffen , and m.p .
rast for putting the data of their computations at our disposal .
we are also grateful to r. hammer for useful remarks on the text .
the work of a.v.g . was supported by the _ deutsche forschungsgemeinschaft _ ( project no .
436 rus 17/56/03 ) and by the russian foundation for basic research ( project nos .
04 - 02 - 16580-a and 07 - 02 - 01094-a ) .
simon , g.w . , brandt , p.n . ,
november , l.j . ,
scharmer , g.b . ,
shine , r.a . : 1994 , in rutten , r.j . ,
schrijver , c.j .
, _ solar surface magnetism , nato advanced science institute _ , vol .
433 , kluwer academic publishers , dordrecht , 261 . | the decrease in the rms contrast of time - averaged images with the averaging time is compared between four datasets : ( 1 ) a series of solar granulation images recorded at la palma in 1993 ; ( 2 ) a series of artificial granulation images obtained in numerical simulations by rieutord _
et al . _
@xcite ; ( 3 ) a similar series computed by steffen and his colleagues ( see wedemeyer _
et al . _ , @xcite ) ; ( 4 ) a random field with some parameters typical of the granulation , constructed by rast @xcite .
in addition , ( 5 ) a sequence of images was obtained from real granulation images using a temporal and spatial shuffling procedure , and the contrast of the average of @xmath0 images from this sequence as a function of @xmath0 is analysed .
the series ( 1 ) of real granulation images exhibits a considerably slower contrast decrease than do both the series ( 3 ) of simulated granulation images and the series ( 4 ) of random fields . starting from some relatively short averaging times
@xmath1 , the behaviour of the contrast in series ( 3 ) and ( 4 ) resembles the @xmath2 statistical law , while the shuffled series ( 5 ) obeys the @xmath3 law from @xmath4 on .
series ( 2 ) demonstrates a peculiarly slow decline of contrast , which could be attributed to particular properties of the boundary conditions used in the simulations .
comparisons between the analysed contrast - variation laws indicate quite definitely that the brightness field of solar granulation contains a long - lived component , which could be associated with locally persistent dark intergranular holes and/or with the presence of quasi - regular structures .
the suggestion that the random field ( 4 ) successfully reproduces the contrast - variation law for the real granulation ( rast , @xcite ) can be dismissed . |
consider a linear differential - algebraic equation ( dae ) : @xmath1 where @xmath2 and @xmath3 are given matrices and @xmath4 , @xmath5 are deterministic noises
. we will be interested in solutions for which @xmath6 are _ a priori _ unknown elements of a bounding set @xmath7 where @xmath8 and @xmath9 are symmetric positive definite matrices .
intuitively , @xmath10 implies that the initial state @xmath11 is bounded and the noise signals @xmath12 have bounded energy , i.e. bounded @xmath13 norm .
the latter is a standard assumption in robust control @xcite . in this paper
we extend the classical results on minimax observers ( see @xcite ) to daes .
namely , we estimate a linear function @xmath14 , @xmath15 of the state trajectory of , based on the output @xmath16 .
since contain uncertain terms @xmath17 , it follows that , even if @xmath18 was invertible , there could be various state trajectories @xmath19 of , and hence several values of @xmath20 , which are consistent with @xmath16 . for the case of rectangular @xmath18
, the dae may have several solutions .
this introduces an additional source of uncertainty .
hence , the best we can do is to construct a function @xmath21 such that * for all * state trajectories @xmath19 of , which are consistent with @xmath16 and correspond to @xmath10 , @xmath20 lies in the interval @xmath22 $ ] , and @xmath23 is as small as possible . a dynamical system whose output is @xmath24 and
whose input is @xmath16 is called a _
minimax observer_. in what follows we take @xmath25 for some @xmath26 .
the reason for this is that the state vector @xmath19 of may contain components which are free , possibly discontinuous , exogenous variables .
the estimation of the latter is problematic even for lti systems @xcite , if the observer is required to be an lti system .
in contrast , @xmath27 is always continuous and it is subject to a differential equation .
our main contribution is a new duality principle ( theorem [ p:5 ] ) .
this result allows us ( i ) to find necessary and sufficient conditions for existence of a minimax observer @xmath24 for finite and infinite time horizons ( theorems [ p:1:theo],[p:5 ] ) , and ( ii ) to show that @xmath24 can be represented as an output of a time - varying linear system in the finite horizon case ( theorem [ t : observer - fin - hor ] ) , and as an output of a stable lti system in the case of infinite time horizon ( theorem [ t : observer - infin - hor ] ) .
finally we prove that for bounded @xmath4 and @xmath5 with possibly unbounded @xmath13 norm , the estimation error remains bounded . the need for estimating the state of a general dae arises in many applications , as it has been demonstrated by various authors , e.g. , @xcite . in this paper
we extend the classical minimax framework to daes of the form .
one motivation is to provide control - theoretic grounds for solving an _ issue of approximation and state estimation for partial differential eeuqations _ ( pdes ) , described in @xcite .
this issue may be resolved by using our results to incorporate pde s approximation error into the estimation process , see section [ sec : numerical - example ] for an example . to the best of our knowledge ,
main results of this paper are new .
its preliminary version appeared in @xcite .
with respect to @xcite the main differences are : ( i ) new ( necessary and sufficient ) conditions for existence of minimax observers ( theorem [ p:5 ] ) , ( ii ) detailed proofs .
duality principle for non - stationary daes was introduced in @xcite , provided @xmath28 .
it was then used to derive a sub - optimal observer .
in contrast , our duality theorems hold for daes with uncertain @xmath29 and the constructed observers are optimal in that the worst - case estimation errors associated with the observers are minimal ( see definitions [ d : finhorobs]-[d : infhorobs ] ) .
the algorithm of @xcite constructing a finite horizon minimax observer by using projectors @xmath30 is a special case of the one of this paper .
minimax observers for discrete time daes were considered in @xcite .
we note that some papers ( see @xcite for an overview ) derive observers for regular daes by converting it to the weierstrass canonical form .
if dae has index @xmath31 , then the resulting ode depends on the derivatives of the noise of order @xmath32 , and so the observer would require bounds for them @xcite .
in contrast , our observer is independent of dae s index and works for bounded @xmath4 and @xmath5 .
the papers @xcite present design of @xmath33-observers which force @xmath33-norm of the `` error system '' to be less than a given number . in contrast , ( i ) minimax observers of this paper minimize a different error measure , ( ii ) unlike @xcite we allow for non - regular daes , and ( iii ) unlike @xcite we present necessary and sufficient conditions for observer existence ( @xmath0-detectability , theorem [ p:5 ] ) .
sufficient existence conditions ( partial impulse observability ) for the asymptotic functional observer for were proposed in @xcite provided @xmath34 .
in contrast , we deal with noisy systems and present necessary and sufficient conditions .
design of unknown input observers for daes was considered in @xcite .
since we deal with bounded unknown inputs and estimate just a part of dae s state , our results do not compare directly with @xcite . in this paper
we consider deterministic noise , stochastic or fuzzy uncertainties have been addressed for example in @xcite .
the definitions of minimax observers are formulated in section [ sec : problem - statement ] .
the main results are presented in section [ sec : main ] .
the application of the observers to pdes is given in section [ sec : numerical - example ] .
the proofs are given in section [ sect : proof ] .
@xmath35 denotes the @xmath36 identity matrix ; for a matrix @xmath37 , @xmath38 denotes its moore - penrose pseudoinverse .
let @xmath39 $ ] , @xmath40 or @xmath41 .
denote by @xmath42 , @xmath43 , @xmath44 respectively the set of all absolutely continuous , the set of all square integrable , and the set of all locally square integrable functions of the form @xmath45 .
recall that @xmath45 is locally square integrable if its restriction to any compact interval @xmath46 is square integrable .
we write @xmath47 , @xmath48 and @xmath49 referring to @xmath42 , @xmath43 and @xmath50 respectively , when @xmath51 is clear from the context .
we say that @xmath52 holds almost everywhere ( a.e . ) on @xmath53 if @xmath54 only on a subset of @xmath53 of measure zero .
we identify linear time invariant ( lti ) systems @xmath55 , @xmath56 , with the tuple of matrices @xmath57 . by convention
, the infimum of an empty set will be @xmath58 .
let @xmath59 $ ] , @xmath60 .
for any @xmath61 , define @xmath62 by @xmath63 a.e .
we begin with a precise definition of dae s solution on the interval @xmath53 of the form @xmath59 $ ] , @xmath64 or @xmath65 .
[ sol : behav ] a _ solution _ of on @xmath53 is a tuple @xmath66 such that : @xmath27 is absolutely continuous and @xmath67 for all @xmath68 and @xmath69 a.e . on @xmath53 .
denote the set of all solutions on @xmath53 by @xmath70 .
let us fix the symmetric positive definite matrices @xmath9 .
a _ solution _ @xmath71 is said to be admissible if @xmath72 for all @xmath73 . denote the set of all admissible solutions by @xmath74 .
[ r : rho_norm_fx ] the map @xmath75 is a norm of the space @xmath76 and @xmath77 is the unit ball of @xmath78 induced by this norm .
next , we define finite horizon minimax observers . fix a time @xmath60 and set @xmath59 $ ] .
fix @xmath79 . to define an observer we rely upon classical results @xcite on minimax ( worst - case ) estimators ( observers ) which state
that for general linear equations the minimax observer is linear in outputs @xmath16 @xcite .
the latter suggests that an estimate @xmath80 of the true value @xmath81 , which is computed by the minimax observer , should be linear in @xmath16 .
it is then natural to model the effect of applying minimax observers to outputs as continuous linear functionals , generated by functions @xmath82 as follows : @xmath83 hence , given @xmath16 , we compute the estimate @xmath80 of the true value @xmath81 by selecting a function @xmath84and evaluating @xmath85 . in order to select @xmath84 optimally we introduce a cost function @xmath86 measuring the worst - case estimation error : @xmath87 [ d : finhorobs]we say that @xmath88 is a _ minimax observer on @xmath59 $ ] _ , if @xmath89 . if @xmath90 , then the estimation error satisfies @xmath91 for * any * admissible solution @xmath92 .
if , in addition , @xmath84 is a minimax observer on @xmath53 , then the error bound @xmath93 is the smallest possible .
we show below that if there exists a function @xmath82 with finite worst - case error , then there also exists the minimax observer @xmath94 , and it can be implemented as an output of a linear system .
consider now the infinite horizon case , i.e. @xmath65 . similarly to the finite horizon case , observers of interest are devices which map outputs @xmath16 to estimates @xmath95 of @xmath96 , and @xmath95 is a linear and continuous function of @xmath97 . to model
such observers we define @xmath98 the set of all maps @xmath99\ } \rightarrow \mathbb{r}^p$ ] such that @xmath100 , the function @xmath101 \ni s
\mapsto u(\tau , s)$ ] belongs to @xmath102,\mathbb{r}^p)$ ] .
an element @xmath103 maps output @xmath104 to a function @xmath105 defined as : @xmath106 now , to construct @xmath95 we take @xmath103 and compute @xmath107 . to select @xmath84 optimally
we define the worst - case estimation error : @xmath108 [ d : infhorobs ] we say that @xmath109 is an _ infinite horizon minimax observer _ , if @xmath110 if @xmath111 , then for large enough @xmath112 , the estimation error satisfies @xmath113 for * any * admissible solution @xmath114 .
if , in addition , @xmath84 is an infinite horizon minimax observer , then the error bound @xmath115 is the smallest possible .
assume that @xmath116 and let @xmath117 denote the minimax observer corresponding to @xmath118 , @xmath119 respectively . since @xmath120 , it follows that @xmath121 is the restriction of @xmath122 onto @xmath123 .
this reveals the intuition behind the definition of the infinite horizon minimax observer @xmath124 , namely note that @xmath124 is a collection of maps @xmath125 such that each map estimates @xmath81 .
this estimate is not optimal , that is @xmath126 , but @xmath127 approaches @xmath128 for @xmath129 as it follows from the representations - given in theorems [ t : observer - fin - hor]-[minimax : obs1 ] .
note that @xmath94 and @xmath130 are unique if they exist and existence of the latter implies existence for the former
. we will be interested in infinite horizon minimax observers which can be represented by stable lti systems .
@xmath103 is said to be represented by an lti system @xmath131 , if for any @xmath104 , where @xmath59 $ ] , @xmath60 or @xmath65 , it holds that @xmath132 , where @xmath133 , @xmath134 .
note that @xmath131 represents an element @xmath84 of @xmath98 , if and only if @xmath135 for all @xmath136 $ ] .
moreover , we will prove that if there exist a @xmath103 with finite worst - case estimation error , then there exist an infinite horizon minimax observer @xmath124 , which , in addition , can be represented by a stable lti system .
we show that the minimax observer is related to the solution of an optimal control problem for the dual dae : @xmath137 to this end , we need to define what we mean by a solution of .
let @xmath53 be an interval of the form @xmath59 $ ] , @xmath64 or @xmath65 .
a pair @xmath138 is a solution of on @xmath53 , if @xmath139 is absolutely continuous and @xmath140 for all @xmath141 .
denote the set of all solutions on @xmath53 by @xmath142 and define @xmath143 .
[ d : l - obs ] the tuple @xmath144 is _
@xmath0-impulse observable _
, if @xmath145)\ne\varnothing$ ] for some @xmath64 .
we define now the following cost function for .
fix @xmath146 and let @xmath53 be an interval of the form @xmath147 $ ] for some @xmath148 , or @xmath65 . for any @xmath149 , define @xmath150 where @xmath151 with @xmath152 and @xmath153 and @xmath154 is defined in .
we can now state the following theorem relating existence of minimax observers to minimizing the cost function .
[ p:1:theo ] the following statements are equivalent : * @xmath144 is @xmath0-impulse observable , * there exists @xmath155,\mathbb{r}^p)$ ] such that @xmath156 , * there exists a minimax observer @xmath94 , and @xmath157 and @xmath158 , where @xmath159)$ ] is a solution of the dual control problem : @xmath160 .
moreover , for any @xmath155,\mathbb{r}^p)$ ] , @xmath161 in particular , the set @xmath162 is not empty , if and only if @xmath156 .
next , we present the notion of @xmath0-detectability and a new duality principle for the infinite horizon case . [
d : l - detect ] the tuple @xmath144 is _
@xmath0-detectable _ , if there exists @xmath163 such that @xmath164 .
note that definition [ d : l - detect ] implies that @xmath145)\ne\varnothing$ ] for some @xmath64 : indeed , for @xmath165 from the definition [ d : l - detect ] it follows that its restriction @xmath166},u|_{[0,t_1]})$ ] onto @xmath167 $ ] belongs to @xmath168)$ ] and hence for any @xmath169 @xmath168 ) \ne \emptyset$ ] .
hence , @xmath0-detectability implies @xmath0-impulse observability by definition [ d : l - obs ] .
[ p:5 ] the following statements are equivalent : * @xmath144 is @xmath0-detectable , * there exists an infinite horizon minimax observer @xmath124 , and @xmath170 , @xmath171 and @xmath172 , where @xmath173 satisfies @xmath174 ) } j(q , u,\tau ) < + \infty\ , .
\end{split}\ ] ] the assumption @xmath175 is essential for theorems [ p:1:theo ] [ p:5 ] .
[ rem_impulse_obs ] note that @xmath144 is @xmath0-impulse observable _ for all @xmath26 _ , if and only if it is impulse controllable in the sense of @xcite .
the latter can be characterized in terms of @xmath144 . if @xmath176 , then @xmath144 is @xmath0-impulse observable for all @xmath26 if and only if @xmath144 is impulse observable in the sense of @xcite . from
* lemma 4.3.7 ) it follows that @xmath144 is @xmath0-detectable for all those @xmath26 for which it is @xmath0-impulse observable @xmath177 @xmath178 = \max_{s \in \mathbb{c } } \mathrm{rank}\left [ \begin{smallmatrix } sf - a \\ h \end{smallmatrix}\right]$ ] .
below we present algorithms for computing minimax observers .
recall from @xcite the notion of lti systems associated with .
[ linassc : def ] let @xmath179 , @xmath180 , @xmath181 , @xmath182 for some @xmath183 .
the lti system @xmath184 is said to be _ associated with _ , if the following conditions hold : 1 .
either @xmath185 and @xmath186=0 $ ] or @xmath187 is full column rank .
2 . let @xmath188 and @xmath189 be the matrices formed by the first @xmath190 rows of @xmath191 and @xmath187 respectively
. then @xmath192 and @xmath193 .
3 . let @xmath53 be an interval of the form @xmath59 $ ] , @xmath64 or @xmath65 .
then @xmath194 if and only if there exists @xmath195 such that @xmath196 a.e . and @xmath197=c_ap+d_ag$ ] a.e .
the matrix @xmath198 is called the state map of @xmath199 .
associated lti systems allow us to represent solutions of daes as outputs of linear systems .
recall from ( * ? ? ?
* theorem 3 - 4 ) that there exists an lti system @xmath184 associated with and that all lti systems which are associated with are feedback equivalent .
+ the matrix @xmath200 allows us to relate the state trajectories of and the state trajectories of @xmath199 : from ( * ? ? ?
* theorem 1 ) it follows that for any @xmath201 , for @xmath202 and @xmath203 -c_av)$ ] , @xmath204 a.e . and
@xmath205 = c_av + d_ag$ ] a.e .
consequently ( see ( * ? ? ?
* corollary 1 ) ) it follows that @xmath144 is @xmath0-impulse observable if and only if @xmath206 . an algorithm for computing the associated lti was discussed in ( * ? ? ?
* proof of th.1 ) and it is summarized in algorithm [ alg1 ] .
this algorithm was implemented , the matlab function @xmath207 implementing algorithm [ alg1 ] is available in the file @xmath208 of the supplemenatry material of this report .
@xmath144 lti system @xmath184 associated with the dual system and the state map @xmath200 .
compute the svd decomposition @xmath209 of @xmath210 and assume that @xmath211 , @xmath212 .
set @xmath213 note that @xmath214 .
let @xmath215 , @xmath216 , @xmath217 , @xmath218 , @xmath219 .
@xmath220 , be such that @xmath221 consider the linear system @xmath222 using the algorithms discussed in ( * ? ? ?
* section 4.5 , page 241 ) , compute a full row rank matrix @xmath223 and a matrix @xmath224 , where @xmath225 is the number of columns of @xmath226 , such that * @xmath227 is the largest output nulling controlled invariant subspace @xmath228 of , and * @xmath229 for instance , @xmath230 can be computed by applying the matlab function @xmath231 from @xcite to @xmath232 .
let @xmath233 be such that @xmath234 , and either @xmath235 , or @xmath236 .
for instance , using matlab functions @xmath237,@xmath238,@xmath239 of @xcite , the function call @xmath240 will return @xmath241 .
define @xmath242 let @xmath243 and @xmath244 be the matrix representations of the linear maps @xmath245 , @xmath246 , @xmath247 with respect to the basis represented by the columns of @xmath223 , which form a basis of @xmath228 , i.e. @xmath248 where @xmath249 is the matrix of the projection operator onto @xmath250 .
let @xmath251 , where @xmath188 is the matrix formed by the first @xmath190 rows of @xmath191 .
now we describe how to construct a minimax observer @xmath252 on the interval @xmath59 $ ] , @xmath253 .
let @xmath184 be an lti system associated with .
consider the equation : @xmath254\ , , \\ & k(t)\!\!=\!\ ! \left\{\!\!\ !
\begin{array}{rl } ( d_a^{\top}sd_a)^{-1}(b_a^{\top}p(t)+d_a^{\top}sc_a ) & \!\!\ ! \mbox { if } d_a\ne 0\ , , \\ 0 \in \mathbb{r}^{1 \times \hat{n } } & \!\!\ !
\mbox { if } d_a = 0\ , . \end{array}\right .
\end{split}\ ] ] where @xmath188 is as in definition [ linassc : def ] .
note that is well defined , as either @xmath187 is full column rank , and so @xmath255 is invertible , or @xmath256 .
let @xmath257 , @xmath258 be such that @xmath259 where @xmath260 and @xmath261 are the matrices formed by the last @xmath262 rows of @xmath191 and @xmath187 respectively , and @xmath263 is as in .
[ t : observer - fin - hor ] if @xmath144 is @xmath0-impulse observable , then @xmath264 is a finite horizon minimax observer on @xmath59 $ ] .
moreover , @xmath265 and @xmath266 , where @xmath267 , @xmath268 and @xmath269 note that @xmath0-impulse observability can easily be checked , see algorithm [ alg5 ] below .
the matlab function @xmath270 implementing algorithm [ alg5 ] is available in the file @xmath208 of supplementary material of this report .
@xmath271 yes , if @xmath144 is @xmath0-impulse observable , no otherwise use algorithm [ alg1 ] to compute an lti system @xmath184 associated with . [
step : check2 ] let @xmath188 be the matrix formed by the first @xmath190 rows of @xmath191 .
check if @xmath0 belongs to @xmath272 , for example , by checking if @xmath273 .
if @xmath274 , return yes , otherwise return no .
theorem [ t : observer - fin - hor ] yields an algorithm for computing the finite horizon minimax estimate .
this algorithm is presented in algorithm [ alg2 ] below .
@xmath271 and the observed output signal @xmath275,\mathbb{r}^p)$ ] minimax estimate @xmath276 use algorithm [ alg1 ] to compute the linear system @xmath184 associated with and the state map @xmath200 .
compute @xmath277 and @xmath263 by solving and find @xmath278 by solving .
the numerical approximations of @xmath277 and @xmath23 may be done by using mbius time integrators @xcite .
return @xmath266 project @xmath0 onto range of @xmath279 and go to 2 .
next , we present an algorithm for computing an infinite horizon minimax observer @xmath252 . to this end ,
recall from @xcite the notion of a stabilizable lti system associated with the dae .
let @xmath184 be an lti system associated with and consider the stabilizability subspace @xmath280 of @xmath199
. from @xcite it then follows that @xmath280 is @xmath281-invariant and @xmath282 .
hence , there exists a basis transformation @xmath283 such that @xmath284 $ ] , @xmath285 and in this new basis , @xmath286 $ ] , @xmath287 $ ] , @xmath288^{\top}$ ] , where @xmath289 , @xmath290 , @xmath291 .
[ assco_stab : def ] the lti system @xmath292 with @xmath293 is called a stabilizable lti system associated with .
the matrix @xmath294 t{\mathcal{m}}$ ] is called the associated state map .
the lti system @xmath295 is a restriction of @xmath199 to the subspace @xmath280 .
it follows that @xmath295 is stabilizable .
moreover , since all associated lti systems of are feedback equivalent , all stabilizable associated lti systems of are feedback equivalent . from the dicussion after lemma 1 of @xcite it follows that the lti @xmath295 and the map @xmath296 can be computed from the matrices @xmath297 . for the convenience of the reader , we present the alghorithm in algorithm [ alg3 ] below .
the matlab function @xmath298 implementing algorithm [ alg3 ] can be found in the file @xmath208 in supplementary material of this report .
@xmath144 stabilizable lti system @xmath292 associated with and the state map @xmath296 .
use algorithm [ alg1 ] to compute the lti system @xmath184 associated with and the state map @xmath200 .
[ step : stable1 ] compute the controllability matrix @xmath299 of @xmath199 , compute a matrix @xmath300 such that @xmath301 is largest @xmath281-invariant subspace such that the restriction of @xmath281 to @xmath301 is stable .
if we use the matlab tool @xcite , then @xmath300 can be taken as the first component of the output of the @xmath302 , [ step : stable2 ] let @xmath303 be any matrix whose columns form a basis of @xmath304 .
then @xmath223 is full column rank , and from ( * ? ? ?
* theorem 4.30 , page 96 ) it follows that @xmath305 equals the largest stabilizable subspace @xmath280 of @xmath199 . in matlab
, @xmath303 can be taken as the output of @xmath306 .
@xmath307 where @xmath308 is the matrix of the projection operator onto @xmath309
. note that by @xcite , @xmath144 is @xmath0-detectable if and only if @xmath310 belongs to the stabilizability subspace @xmath280 of @xmath199 .
this remark yields the following algorithm , presented in algorithm [ alg6 ] , for checking @xmath0-detectability .
@xmath271 yes , if @xmath144 is @xmath0-detectable , no otherwise use algorithm [ alg2 ] to compute an lti system @xmath184 associated with and the state - map @xmath200 . use steps [ step : stable1]-[step : stable2 ] of algorithm [ alg3 ] to compute a full column rank matrix @xmath303 such that @xmath305 is the largest stabilizable subspace @xmath280 of @xmath199 like in step [ step : check2 ] of algorithm [ alg5 ] , check if @xmath274 , where @xmath188 is the matrix formed by the first @xmath190 rows of @xmath191 .
if @xmath311 , then quit the algorithm and return no .
if @xmath274 , then check if @xmath312 .
the latter can be done , for example , by checking if @xmath313 , where@xmath308 ) if @xmath314 , then return yes , otherwise return no .
the matlab function @xmath315 implementing algorithm [ alg6 ] can be found in the file @xmath208 of the supplementary material of this report .
let @xmath292 and @xmath296 be as in definition [ assco_stab : def ] . consider the following algebraic riccati equation : @xmath316\ , .
\end{split}\ ] ] note that either @xmath293 is full column rank , and so , by positive definiteness of @xmath317 , @xmath318 is invertible , or @xmath319 . if @xmath320 is a solution of , then let @xmath321 , @xmath322 be such that @xmath323 where @xmath324 and @xmath325 are the matrices formed by the last @xmath262 rows of @xmath326 and @xmath327 respectively .
define @xmath328 [ t : observer - infin - hor ] [ main : obs ] assume that @xmath144 is @xmath0-detectable .
then the following holds : * @xmath329 is a hurwitz matrix and has a unique symmetric positive definite solution @xmath277 , * @xmath109 , where @xmath330 , @xmath331 , is an infinite horizon minimax observer . *
the minimax error is @xmath332 , * @xmath124 can be represented by the lti system @xmath131 , i.e. @xmath333 where @xmath334 solves @xmath335 we stress that one does not need to solve several times in order to compute the minimax observer @xmath336 for various @xmath337 , @xmath338 . indeed ,
@xmath329 and @xmath339 do not depend on @xmath0 , since they depend only on the matrices @xmath340 of the associated stabilizable lti @xmath295 and on the weight matrices @xmath341 and @xmath299 . in turn , @xmath295 depends only on the matrices @xmath297 , see algorithm [ alg3 ] .
hence , it is sufficient to define @xmath342 where @xmath343 $ ] , provided @xmath144 is @xmath337-detectable , @xmath338 .
[ obs : error ] assume @xmath144 is @xmath0 detectable and let @xmath124 be the infinite horizon minimax observer . for any @xmath114 , the estimation error @xmath344 satisfies @xmath345\ , , \\ & \widetilde{r}(0)=(c_g - d_gk)^{\top } \left[\begin{smallmatrix } fx(0 )
\\ 0 \end{smallmatrix}\right ] \ , , \\ & e(t):=\ell^{\top}fx(t)-{\mathscr{o}_{\widecheck u}(y)(t)}=\ell^{\top}f{\mathcal{m}}^{\top}_g \widetilde{r}(t)\ , . \end{split}\ ] ] in particular , if @xmath34 , then @xmath346 . if @xmath17 are bounded , i.e. @xmath347 , then the estimation error @xmath348 is bounded , i.e. @xmath349 . if @xmath350 , @xmath351 , then @xmath352 .
this lemma reveals the following important points : ( i ) the infinite horizon minimax observer behaves like an asymptotic observer in the absence of noises ; ( ii ) if the noise signals are of bounded energy ( i.e. they are in @xmath353 ) , then the estimation error is of bounded energy ; ( iii ) if the noise signals are just bounded , then the estimation error is bounded too , i.e. , the upper bound of the error is proportional to the upper bound of the noises .
the discussion above allows us to formulate an algorithm for computing an infinite horizon state estimate ( see algorithm [ alg4 ] ) . the matlab function @xmath354 in the file @xmath208 in the supplementary material of this report implements steps [ alg4:step1 ] [ alg4:step2 ] algorithm [ alg4 ] .
@xmath271 and the observed output @xmath355 infinite horizon minimax state estimate @xmath356 .
[ alg4:step1 ] use algorithm [ alg3 ] to compute the stabilizable lti system @xmath292 associated with and the state map @xmath296 . compute the solution @xmath277 of and compute the feedback matrix @xmath263 from . in matlab
, the command @xmath357 can be used to calculate @xmath277 and @xmath263 .
[ alg4:step2 ] set @xmath358 solve the differential equation to find @xmath278 . return @xmath359
consider the following infinite - dimensional system with outputs , determined by the partial differential equation ( pde ) of heat transfer : @xmath360 where @xmath361 and @xmath362 take values in the space @xmath363,\mathbb{r})$ ] , i.e. , @xmath361 is itself a function from @xmath364 $ ] to @xmath365 , and @xmath366 for all @xmath367 such that @xmath368 .
moreover , @xmath369 , @xmath370 is given by : @xmath371 for some @xmath372,\mathbb{r}^p)$ ] , @xmath373 .
we say that a pair @xmath374 solves if @xmath375 is a frechet differentiable function such that @xmath376 is defined , and holds a.e . , and @xmath377 .
physically , @xmath378 represents the temperature at time @xmath379 at point @xmath380 $ ] .
the constant @xmath381 in the definition of the operator @xmath382 is the thermal diffusivity , in this example it is taken to be @xmath383 @xmath384 . from the control perspective
, @xmath385 is a state , @xmath386 are noises and @xmath16 is a measured output .
we take @xmath387 and @xmath388 , for other choices of @xmath389 and @xmath262 the discussion is similar . based on @xmath16 and assuming that @xmath390 and @xmath391 satisfy @xmath392 , @xmath393 denotes the standard norm on @xmath394 , we would like to estimate @xmath361 from @xmath16 .
we recast the problem of estimating @xmath385 as the minimax estimation problem for daes as follows ( see @xcite for the details ) : consider the basis @xmath395 of @xmath394 , where @xmath396 , @xmath397 , with @xmath398 denoting the @xmath399th legendre polynomial , @xmath400 .
fix an integer @xmath401 and define @xmath402 where @xmath403 is the standard inner product in @xmath394 and @xmath404 is the kronecker delta symbol for all @xmath405 .
consider the dae with @xmath406 assume @xmath374 solve .
from @xcite it follows that the tuple @xmath407 is a solution of , provided @xmath408 , @xmath409 , @xmath410 , and @xmath411 , and @xmath412 , where @xmath413 , @xmath414 for some @xmath415 , @xmath416 .
moreover , if @xmath385 and its partial derivatives are bounded ( which is the case for physically meaningful solutions ) , then by @xcite , there exist @xmath417 , @xmath418 such that @xmath419 .
the intuition behind this is as follows : the component @xmath420 of @xmath421 represents the first @xmath422 coordinates of @xmath361 w.r.t the basis @xmath423 , and @xmath424 models the impact of @xmath425 , the error of approximating @xmath385 by its projection onto the first @xmath422 basis functions @xmath426 , onto @xmath420 .
for this reason , @xmath424 is included into the state of dae .
the latter allows to incorporate the projection error @xmath427 into the estimation process .
the smaller @xmath424 is , the closer the behavior of is to that of .
the algebraic equations @xmath428 and @xmath429e_m+f_3=0 $ ] together with @xmath430 imply that the @xmath13 norm of @xmath431 and of the first @xmath432 components of @xmath424 is uniformly bounded , i.e. these functions live in an ellipsoid .
note that the expressions for @xmath317 provided by ( * ? ? ?
* eq . ( 3.12 ) and ( 3.13 ) )
are very conservative , and , in practice , @xmath317 are found by tuning .
the minimax observer for can be used to estimate the state of as follows .
assume that is @xmath0-detectable for @xmath433 , where @xmath434 is the @xmath399th standard basis vector in @xmath435 .
let @xmath436 denote the minimax observer corresponding to @xmath437 according to theorem [ main : obs ] , and let @xmath438 .
define @xmath439 .
if @xmath374 is a solution of , then , as noted above , the tuple @xmath407 solves , and @xmath440 , the projection of @xmath385 to the first @xmath432 basis functions . by noting that @xmath441 is the minimax estimate of @xmath442
, we derive that @xmath443 is the minimax estimate of the vector @xmath444 .
this latter estimate is `` good '' if the minimax error @xmath445 is `` small '' . for the simulations we chose @xmath446 , @xmath447 , @xmath448 , @xmath449 .
note that the entries of @xmath450 , and @xmath451 can be computed analytically ( see ( * ? ? ?
* example 7.2 , p. 121 ) ) so that @xmath452 can be computed numerically .
the numerical values of the matrices @xmath297 can be found in the file @xmath453 ( matlab format ) or in the files @xmath454 , @xmath455 , @xmath456 ( csv text formal ) of the supplementary material of this report .
if we check @xmath457 , then it follows that its rank is less than its number of columns .
hence , by remark [ rem_impulse_obs ] , there exists @xmath0 such that @xmath144 is not @xmath0-impulse observable for this @xmath0 .
thus , the observer proposed in @xcite is inapplicable as it requires to solve a linear equation ( * ? ? ?
* eq . ( 5c ) ) which has no solution if there exists an @xmath0 for which @xmath144 is not @xmath0-impulse observable . note that @xcite is not applicable either : by ( * ? ? ?
* remark 1 ) , existence of a filter in the sense of @xcite implies that the dae at hand is regular , which is not the case .
however , using algorithm [ alg5 ] - [ alg6 ] , it can be checked numerically that @xmath144 is @xmath0-impulse observable and @xmath0-detectable ( def.[d : l - obs]-[d : l - detect ] ) only for @xmath458 , @xmath459 , i.e we can estimate ( with finite worst - case error ! ) only the first @xmath432 components of @xmath27 . to generate the outputs we took an input @xmath460 with @xmath461 and @xmath462 , initial condition @xmath463 , and noise @xmath464 .
then we computed the exact solution @xmath465 of by using eigen - basis of the operator @xmath382 , i.e. , @xmath466 and @xmath467 , @xmath468 , @xmath469 .
as noted , @xmath465 yields a solution of , namely @xmath470 and @xmath471 .
we applied algorithm [ alg6 ] to check that the dae at hand is @xmath0-detectable for all standard unit vectors of @xmath472 .
we used the matlab function @xmath315 from the file @xmath208 in the supplementary material of this report to check @xmath0-detectability .
we then applied algorithm [ alg4 ] to compute infinite horizon minimax observer @xmath473 for to estimate @xmath474 on @xmath475 $ ] . to this end
we set @xmath476 , @xmath477 .
we used the matlab function @xmath478 from the file @xmath208 of the supplementary material to calculate the matrices @xmath479 of the minimax observer .
these matrices can be found in the files @xmath480 ( csv text format ) and @xmath481 ( matlab data file format ) included in the supplementary material of this report .
the results of the intermediate steps of the computation can also be found in the supplementary material .
more precisely the matrices of the lti @xmath184 associated with the dual system and the state map @xmath200 can be found in the csv text files @xmath482 , @xmath483 , @xmath484 , @xmath485 , @xmath486 of the supplementary material of this report .
the matrices of the stabilizable associated lti @xmath292 and the state map @xmath296 can be found in the csv text files @xmath487 , @xmath488 , @xmath489 , @xmath490 , @xmath491 of the supplementary material of this report .
the solution @xmath277 of and the gain @xmath263 can be found in the csv text files @xmath492 and @xmath493 of the supplementary material .
it turns out that @xmath473 is quite close to @xmath474 component - wise .
specifically , figure [ fig ] shows that the minimax estimate @xmath494 tracks the components @xmath495 , @xmath496 of dae s state vector quite well .
note that @xmath474 represents the coordinates of the true solution @xmath497 with respect to the basis @xmath395 .
hence , it does not itself represent a physical quantity and therefore it has no measurement unit .
we also compared @xmath443 , the minimax estimate of the vector @xmath498 against @xmath499 , and they were in a good agreement .
in particular , on figure [ fig ] we present the plots of @xmath500 versus @xmath501 for @xmath502 .
note that @xmath503 represents an estimate of a physical quantity , namely , the temperature at time @xmath112 at point @xmath504 . versus @xmath505 , @xmath496 and @xmath506 versus @xmath503 for @xmath502 .
let us prove that the set @xmath162 is not empty , if and only if @xmath156 .
this will also prove @xmath507 .
take @xmath508 such that @xmath509 is absolutely continuous and @xmath510 .
define @xmath511 . clearly , @xmath512 .
now , let @xmath513 . then , integrating by parts ( see ( * ? ? ?
* f.(2.1 ) ) ) , we find : @xmath514 let @xmath241 denote a differential operator associated to : @xmath515 with domain @xmath516 .
define the adjoint of @xmath241 , namely @xmath517 , where the domain @xmath518 of @xmath519 is given by : @xmath520 now we claim that @xmath521 for some @xmath522 .
note that equality @xmath523 implies that , by definition of @xmath519 and @xmath391 , @xmath524 verifies @xmath525 and @xmath526 , so that @xmath527 , and vice versa , if @xmath527 then @xmath528 , @xmath529 verifies @xmath523 .
thus , it suffices to prove the above claim . to this end , recall the definition of @xmath530 ( remark [ r : rho_norm_fx ] ) .
we note that @xmath531 and so @xmath532 .
thus , by and , we have that @xmath533 , where @xmath534 to compute @xmath535 we recall from convex analysis @xcite that the support function @xmath536 of the ellipsoid @xmath537 equals to @xmath538 .
analogously , the support functional @xmath539 of the ellipsoid @xmath540 equals to @xmath541 .
now we note that @xmath535 equals to the support function of the pre - image of @xmath540 with respect to the operator @xmath542 .
the latter can be computed by using young - fenhel conjugate ( see ( * ? ? ?
* f.(2.5),lemma 2.2 ) ) : @xmath543=\bigl[\begin{smallmatrix } b(w , u ) \\
u \end{smallmatrix}\bigr]\}\ , . \end{split}\ ] ] this equality implies that @xmath544 for some @xmath522 .
let us now prove . if @xmath545 , then from the discussion above both sides of are @xmath58 .
assume @xmath546 .
take @xmath547 such that @xmath548 .
set @xmath549 .
then @xmath550 in and so we get : @xmath551 where @xmath552 , @xmath553 is the set of all @xmath554 such that @xmath555 for some @xmath556 , and @xmath557 is the support function of @xmath553 .
thus , to compute @xmath93 it suffices to compute right hand side of . from
* f.(2.6),lemma 2.2 ) : @xmath558 where @xmath559 .
we claim that the right - hand side of equals to @xmath560 .
indeed , since @xmath561 , @xmath562 , @xmath153 it follows that we can fix @xmath389 and eliminate @xmath563 by minimizing the 1st term of @xmath154 w.r.t .
we get : @xmath564 where @xmath565 . now , to prove the above claim we note that @xmath566 . in other words , if @xmath567 then @xmath568 is a sum of @xmath569 and @xmath570 provided @xmath571 .
this allows us to write : @xmath572 = \inf_{(q , u ) \in \mathscr{d}_{\ell}(i ) , u=\delta_{t_1}(u ) } [ \\ & ( f^{\top}q(t_1))^{\top}\bar q_0f^{\top}q(t_1 ) + \int_0^{t_1}q^{\top}(t)q^{-1}q(t)dt]\ , , \end{split}\ ] ] were @xmath573 .
thus , the right - hand side of equals @xmath560 .
the latter and prove . to prove @xmath574 we note
that , by definition [ d : finhorobs ] , @xmath575 implies that @xmath576 and then @xmath577 follows .
now , assume that @xmath577 holds and @xmath578 denotes a minimizer of @xmath579 over the affine set @xmath580 .
then , by , @xmath157 is a minimizer for @xmath93 and so @xmath575 holds by definition [ d : finhorobs ] . to prove @xmath581 we assume that @xmath144 is @xmath0-detectable . then is behaviorally stabilizable from @xmath0 ( according to the terminology of @xcite ) and vice - versa .
then by ( * ? ? ?
* theorem 6 ) there exists @xmath582 such that holds .
let us prove that @xmath330 verifies . to this end , from it follows that for any @xmath583,\mathbb{r}^p)$ ] : @xmath584 ) } j(q , v,\tau)\ , .
\end{split}\ ] ] take any @xmath103 . then by using and
and we get : @xmath585 ) } j(q , v,\tau ) = j^*_\infty\ , .
\end{split}\ ] ] apply to @xmath586 .
then @xmath587 ) , v={u^*}|_{[0,\tau ] } } j(q , v,\tau)\le j(q^*,u^*,\tau)$ ] , so that @xmath588 .
this and implies that @xmath589 . since @xmath590 by it follows that @xmath591 and so @xmath130 verifies . to prove @xmath592
let @xmath109 be an infinite horizon minimax observer .
then , by definition [ d : finhorobs ] , @xmath593 .
hence , there exists @xmath594 such that for all @xmath595 : @xmath596 .
now , by of theorem [ p:1:theo ] we get that @xmath597 ) , u=\delta_{t}(\widecheck{u}(t,\cdot ) ) } j(q , u , t)$ ] for all @xmath595 , and so @xmath598 ) , u=\delta_{t}(\widecheck{u}(t,\cdot ) ) } j(q , u , t)<+\infty$ ] .
the latter and ( * ? ? ?
* theorem 6 ) implies that is behaviorally stabilizable from @xmath0 that , as noted above , implies @xmath599 .
assume that @xmath600 are defined by .
set @xmath601 , @xmath602 . from (
* theorem 5 ) it then follows that @xmath603 and @xmath604 . from theorem [ p:1:theo ]
it follows that @xmath252 is indeed the minimax observer on @xmath53 and @xmath605 finally , let @xmath23 be the solution of .
then @xmath606 , and by integrating both sides , @xmath607 .
if @xmath144 is @xmath0-detectable , then is behaviorally stabilizable from @xmath0 ( according to the terminology of @xcite ) and vice - versa .
hence , the statement of @xmath599 follows from ( * ? ? ?
* lemma 2 ) . from
* theorem 6 ) it follows that if @xmath322 is a solution of @xmath608 and @xmath609 are such that @xmath610=(c_g - d_gk)v^*$ ] , then @xmath582 and @xmath611 ) } j(q , u , t)$ ] and @xmath612 . from @xmath577 , theorem [ p:5 ] it then follows that @xmath613 hold true .
now , to prove @xmath614 we note that from the above discussion @xmath615 and thus @xmath616 represents @xmath124 .
let @xmath617 be such that @xmath618 , @xmath619 , @xmath620 .
define @xmath621 \mapsto ( fx)^\top(s)(f^\top)^{+}f^\top q^*(t - s)=x^\top(s)f^\top q^*(t)$ ] .
since @xmath622 and @xmath27 are both absolutely continuous , @xmath389 is absolutely continuous .
note that @xmath623 is in the range of @xmath18 and so @xmath624 .
hence , @xmath625 integrate both sides of the above equality and use @xmath626 : @xmath627\left[\begin{smallmatrix } f(s ) \\
-\eta(s ) \end{smallmatrix}\right]ds\,.\ ] ] note that @xmath628 , and hence the left - hand side of is the estimation error @xmath348 .
it is easy to see that the right - hand side of equals @xmath629 where @xmath630 solves .
we have presented an original solution of the minimax observer design problem for linear daes .
it is `` application friendly '' as ( i ) it relies upon well developed tools ( e.g. riccati equations ) , ( ii ) it imposes easy to check necessary and sufficient conditions ( @xmath0-detectability ) on @xmath144 and ( iii ) it is optimal for @xmath13 inputs and provides bounded estimation error for merely bounded inputs .
t. berger and t. reis , `` controllability of linear differential - algebraic systems : a survey , '' in _ surveys in differential - algebraic equations i _
, a. ilchmann and t. reis , eds.1em plus 0.5em minus 0.4em springer , 2013 , pp .
161 .
s. zhuk , `` kalman duality principle for a class of ill - posed minimax control problems with linear differential - algebraic constraints , '' _ applied mathematics & optimization _ , vol . 68 , no . 2 , pp . 289309 , 2013 .
m. petreczky and s. zhuk , `` solutions of differential - algebraic equations as outputs of lti systems : application to lq control problems , '' _ conditionally accepted to automatica .
arxive 1312.7547 _ , 2014 . | in this note we construct minimax observers for linear stationary daes with bounded uncertain inputs , given noisy measurements .
we prove a new duality principle and show that a finite ( infinite ) horizon minimax observer exists if and only if the dae is @xmath0-impulse observable ( @xmath0-detectable ) .
remarkably , the regularity of the dae is not required . |
at what epoch did the first generation of galaxies and quasars form
? how was the universe re - ionized , ending the `` dark ages '' ( @xcite ) ?
these fundamental questions can only be answered with studies of high - redshift objects .
the last few years have witnessed the first direct observations of galaxies at redshift higher than five ( @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , see also the review by @xcite ) , while detailed studies of the ensemble properties and large scale distribution of galaxies at @xmath17 have begun ( steidel et al .
1998 , 1999 ) .
several quasars have been found at @xmath18 ( fan et al . 1999 , 2000a , @xcite ) , including a low - luminosity quasar at @xmath19 ( @xcite ) .
studies of high - redshift quasars provide important probes of this critical epoch in cosmic evolution .
the lack of the gunn - peterson ( 1965 ) effect in the absorption spectrum of a @xmath20 quasar ( @xcite ) indicates that the universe is already highly ionized at that redshift .
the exact epoch of re - ionization could be determined from the absorption spectra of quasars at even higher redshift ( @xcite , @xcite ) .
the study of the luminosity function of high - redshift quasars will constrain models of quasar and galaxy evolution ( @xcite , @xcite ) , and determine whether it was uv radiation from agns or from young massive stars that re - ionized the universe , ending the `` dark ages '' ( @xcite ) .
measurements of the chemical abundance in the quasar environment will reveal the metal production process at the very early stage of galaxy evolution ( @xcite ) .
finally , luminous high - redshift quasars represent high peaks of density fields , and may be the markers of large scale structure at these early epochs ( @xcite , @xcite , @xcite ) . the sloan digital sky survey ( sdss ; @xcite ) is using a dedicated 2.5 m telescope and a large format ccd camera ( @xcite ) at the apache point observatory in new mexico to obtain images in five broad bands ( @xmath21 , @xmath22 , @xmath23 , @xmath1 and @xmath24 , centered at 3540 , 4770 , 6230 , 7630 and 9130 , respectively ; @xcite ) over 10,000 deg@xmath16 of high galactic latitude sky .
the multicolor data from sdss have proven to be very effective in selecting high - redshift quasars : more than 50 quasars at @xmath25 have been discovered to date from about 600 deg@xmath16 of imaging data ( fan et al .
1999 , 2000a , @xcite , @xcite ) .
the inclusion of the reddest band , @xmath24 , in principle enables the detection of quasars up to @xmath26 in sdss data . in this paper , we report the discovery of sdssp j104433.04012502.2 ( the name reflecting its j2000 coordinates from the preliminary sdss astrometry , accurate to @xmath27 in each coordinate ) , a very luminous , `` @xmath1-dropout '' quasar at @xmath0 , selected by its very red @xmath28 color . in a @xmath29-dominated flat universe (
@xmath30 = 65 km s@xmath6 mpc@xmath6 , @xmath31 and @xmath32 , referred to as the @xmath29-model in this paper , @xcite , @xcite ) , @xmath0 corresponds to an age of 0.9 gyr in an universe 13.9 gyr old , or a look - back time of 93.2% of the age of the universe .
similarly , the universe was 0.7 gyr old at @xmath0 in a universe 13.0 gyr old at present for a model with @xmath33 and @xmath30 = 50 km s@xmath6 mpc@xmath6 , which we refer to as the @xmath33 model in this paper .
we present the photometric observations and target selection in 2 , and the spectroscopic observations in 3 . in 4 , we discuss the cosmological implications , including the constraints on the gunn - peterson effect , quasar evolution models , and black hole formation .
the object sdssp j104433.04012502.2 ( hereafter sdss 10440125 for brevity ) was selected from the sdss imaging data based on its extremely red @xmath34 color .
the photometry of this object is summarized in table 1 .
the photometric observations of this region were obtained by the sdss imaging camera on 2000 march 4 during the sdss commissioning phase . for off - equator scans ,
the telescope moves along a great circle , with the photometric camera drift - scanning at the sidereal rate .
the effective exposure time is 54.1 seconds in each band .
the seeing in the @xmath1 and @xmath24 bands was about @xmath35 .
the photometric calibration is provided by an auxiliary 20-inch telescope at the same site ( uomoto et al .
2000 , in preparation ) .
the photometric zeropoint is accurate to about 7% in @xmath21 and @xmath24 and 2% in @xmath22 , @xmath23 and @xmath1 . because the definition of the photometric system is not yet finalized , we quote measured magnitudes using asterisks ( e.g. , @xmath36 ) to represent preliminary photometry , while referring to the filters with primes ( e.g. , @xmath1 ) .
sdss 10440125 is undetected in @xmath21 , @xmath22 and @xmath23 .
the object is detected at the 5-@xmath37 level in the @xmath1 band .
the @xmath24 detection is of very high significance , with @xmath38 .
data are quoted as asinh magnitudes ( @xcite ) and are on the ab magnitude system ( @xcite ) .
finding charts for sdss 10440125 in the @xmath1 and @xmath24 bands are shown in figure 1 .
as shown by @xcite , the colors of high - redshift quasars are strong functions of redshift in the sdss filter system , as first the ly@xmath8 forest and then the lyman limit systems move through the filter system . at @xmath39 , quasars become very red in @xmath40 while remaining blue in @xmath41 , and can be readily distinguished from stars based on these colors . at @xmath42 , quasars become very red in @xmath41 .
they are easily distinguished from red stars in the @xmath41 vs. @xmath28 color - color diagram ( fan et al . 1999 , 2000a ) .
quasars at @xmath43 have very similar @xmath41 and @xmath28 colors to those of late m stars .
finally , at @xmath44 , the ly@xmath8 emission line begins to move out of the sdss @xmath1 filter . with a predicted @xmath45 ,
quasars at such redshifts become @xmath1-band dropout objects . however , unlike the lower - redshift @xmath22 and @xmath23-dropout quasars , these quasars have only one measurable optical color , as they will be completely undetected in @xmath23
. the sdss alone can not provide a constraint on the quasar s continuum shape redward of the ly@xmath8 emission .
it is thus difficult to distinguish them from other classes of red objects without additional information , such as near - infrared photometry ( see also zheng et al .
2000 ) or detection in the radio or x - ray .
in fact , the only other known class of stellar objects with such red colors ( @xmath45 ) are the extremely cool stars / substellar objects with spectral type l or t ( @xcite , @xcite , @xcite , @xcite , @xcite ) .
most of these objects are brown dwarfs with mass below the hydrogen burning limit .
although l and t dwarfs are very rare on the sky ( the density for l dwarfs is 1 per @xmath46 deg@xmath16 for @xmath47 , @xcite ) , they are still more numerous than are @xmath48 quasars , and are the major contaminants of searches for @xmath1-dropout quasars . near infrared photometry
can be used to separate high - redshift quasar candidates from these cool dwarfs : the continuum shape of quasars is relatively flat towards near - ir bands , while the flux of l dwarfs continues to rise sharply towards longer wavelengths .
we have selected @xmath1-dropout candidates with @xmath45 and @xmath49 from about 600 deg@xmath16 of sdss imaging data .
these 600 deg@xmath16 overlap the publicly released area of the two micron all sky survey ( 2mass ) .
sdss 10440125 is the only @xmath1-dropout source in the 2mass covered area that is _ not detected _ at the 7-@xmath37 level in any band in the 2mass point source catalog .
all other sources are detected in at least one band at more than 10 @xmath37 ( e.g. @xcite , @xcite ) .
is the non - detection in 2mass sufficient to rule out the possibility that sdss 10440125 is a brown dwarf ? in this area of the sky , the 2mass point source catalog has 7-@xmath37 limiting magnitudes of roughly @xmath50 , @xmath51 and @xmath52 .
this indicates a 7-@xmath37 upper limit on the optical - ir colors of sdss 10440125 : @xmath53 , @xmath54 and @xmath55 .
there are three l dwarfs in fan et al .
( 2000b ) that have @xmath56 : sdss 03300025 ( @xmath57 , spectral type l2 ) , sdss 05390059 ( @xmath58 , l5 ) and sdss 13260038 ( @xmath59 , l8 ) .
they all have @xmath60 , @xmath61 and @xmath62 ( see tables 2 and 5 in @xcite ) .
although the current l dwarf sample is small , the @xmath34 and the @xmath63 ( or @xmath64 , @xmath65 ) colors seem to be reasonable indicators of the spectral type for l dwarfs ( @xcite , @xcite ) therefore , if sdss 10440125 ( @xmath66 ) had been an l dwarf , we would expect it to have been detected by 2mass with high significance .
thus the non - detection in the 2mass passbands indicates that sdss 10440125 is unlikely to be a l dwarf , but is rather a source with a much flatter ir continuum , such as a quasar at @xmath67 or a compact galaxy at @xmath68 .
a @xmath69-band image of sdss 1044 - 0125 was obtained on the night of 2000 april 17 , using the near infrared camera ( nirc , matthews & soifer 1994 ) on the keck i telescope under photometric skies and good seeing .
the observations consisted of a nine - point dither pattern , integrating for 3 @xmath70 10s coadds at each location .
the data were flattened , sky - subtracted , shifted , and stacked using the dimsum package in iraf .
the images show sdss 1044 - 0125 to be an unresolved point source with fwhm = @xmath71 .
photometry through a 2 radius aperture yields @xmath72 , referenced to the standards of persson et al .
there are no companions or associated structure within 20 of the quasar to a @xmath73 point source limiting magnitude of @xmath74 .
low and high dispersion spectra of sdss 10440125 were obtained with the echelle spectrograph and imager ( esi ; @xcite ) on the keck ii telescope on the night of 2000 april 6 .
the night was photometric with @xmath75 seeing .
a 1200 second exposure was taken through a @xmath76 slit in the low - dispersion mode of esi . in this mode ,
a prism is used for dispersion .
the spectrum covers 3900 to about 10000 .
the dispersion varies roughly linearly with wavelength from 0.8 / pixel at 3900 to 10 / pixel at 10000 .
in addition , two 1200 second high resolution spectra were taken through a @xmath77 slit in the echellette mode of esi . in this mode , the spectral range of 3900 to 11000 is covered in ten spectral orders with a constant dispersion of 11.4 km s@xmath6 pixel@xmath6 .
wavelength calibrations were performed with observations of hg - ne - xe lamps in the low - dispersion prism mode , and a cu - ar lamp in the echellette mode .
the spectrophotometric standard g191-b2b ( @xcite , @xcite ) was observed for flux calibration .
all observations were carried out at the parallactic angle .
the data were reduced with standard iraf routines .
figure 2 shows the final spectrum combining the low and high dispersion observations ( total exposure time of 3600s ) , over a wavelength range of 4500 - 10000 , and binned to 4 / pixel .
the telluric absorption bands were removed using the standard star spectrum .
the absolute flux scale of the spectrum is adjusted so that it reproduces the sdss @xmath78 magnitude .
the signal - to - noise ratio at @xmath79 is 15 20 per pixel .
the spectrum of sdss 10440125 shows the unambiguous signature of a very high redshift quasar : the broad and strong ly@xmath8+nv emission line at @xmath80 with a sharp discontinuity to the blue side , due to the onset of very strong ly@xmath8 forest absorption .
the flux level drops by a factor of @xmath81 from the red side to the blue side of the ly@xmath8+nv emission .
a ly@xmath82+ovi emission line is detected at @xmath83 7000 , with an additional flux decrement due to the onset of ly@xmath82 forest absorption lines .
the spectrum shows no detectable flux at @xmath84 because of the presence of a lyman limit system ( see 4 ) .
redward of ly@xmath8 , two additional emission lines , oi+siii@xmath85 and siiv+oiv]@xmath86 are also clearly visible .
the synthetic @xmath28 color calculated from the spectrum in figure 2 is 2.5 , consistent with the sdss measurement .
the redshift determination of such an object is not straightforward .
the ly@xmath8 emission line is severely affected by the strong ly@xmath8 forest lines ( as well as by possible internal absorption lines ) . for @xmath87 quasars , schneider , schmidt & gunn ( 1991 ) show that the peak of the ly@xmath8 emission line is typically at rest - frame 1219 , but with a large scatter .
another strong line , civ@xmath88 1549 , is out of the range of our spectrum . oi+siii@xmath85 and siiv+oiv]@xmath86 are relatively weak .
we use the central wavelengths of these two blends from @xcite .
gaussian fits of these two lines are listed in table 2 .
based on these fits , we adopt a redshift of 5.80 @xmath89 0.02 for sdss 10440125 .
objective measurement of the equivalent width of the ly@xmath8 line is difficult ( see the discussion by schneider , schmidt & gunn 1991 ) . using the continuum level determined from the red side of ly@xmath8 emission , we find the rest - frame equivalent width of ly@xmath8 + nv is @xmath90 . however , this number is clearly an underestimate due to the extreme ly@xmath8 absorption , maybe by a factor of two .
figure 2 indicates that the expected line center of ly@xmath8 at @xmath0 is severely absorbed , and that the blue wing of the line has almost completely disappeared .
the measurement of the ly@xmath82+ovi line is difficult as well .
we smooth the spectrum of figure 2 to estimate the continuum level , and find an equivalent width of @xmath91 .
this value is also highly uncertain due to the difficulty in defining a `` continuum '' in the ly@xmath8 forest region .
it is interesting to note that the strengths of the oi+siii@xmath85 and siiv+oiv]@xmath86 line blends are comparable to those at much lower redshift : the average rest frame equivalent widths of these two lines are @xmath92 and @xmath93 in the sample of 30 quasars at @xmath94 in schneider , schmidt & gunn ( 1991 ) , compared to 3.4 and 7.0 in the case of sdss 10440125 .
previous studies have shown that quasar environments at @xmath94 have roughly solar or higher metallicities ( @xcite ) .
although we can not derive the metallicity of sdss 10440125 based on these emission line measurements , the existence of strong lines suggests that in this system , the metallicity is already quite high at this redshift . assuming the metals were produced in stellar evolution , the initial starburst and chemical enrichment of the quasar environment must have happened at a very early epoch . from the spectrum in figure 2 ,
we derive its continuum ab magnitude at rest frame 1280 , @xmath95 , after correcting for interstellar extinction ( @xmath96 , from the map of @xcite ) . in the @xmath29-model ( see 1 ) , sdss 10440125 has an absolute magnitude @xmath97 . assuming a power law continuum with @xmath98 we find @xmath99 and @xmath100 . in the @xmath33 model ( 1 )
it has @xmath101 , @xmath102 , and @xmath103 . in this cosmology
, the nearby luminous quasar 3c273 has @xmath104 .
sdss 10440125 is a very luminous quasar , about twice as luminous as 3c 273 ( assuming that it is not amplified by lensing or beaming ) .
sdss 10440125 is detected neither in the first radio survey ( @xcite ) at the 1mjy level at 20 cm wavelength , nor in the rosat all sky survey ( @xcite ) , implying a 3-@xmath37 upper limit of @xmath105 ergs @xmath106 s@xmath6 in the 0.1 2.4 kev band .
these result is not unexpected ; only a few @xmath107 quasars have observed x - ray or radio fluxes above this value ( @xcite , @xcite ) .
a deep exposure with chandra or xmm is needed to determine its x - ray properties .
the high luminosity of sdss 10440125 makes it an ideal object for high signal - to - noise ratio observations to study the intergalactic medium at high redshift . in order to detect continuum break caused by the lyman limit system ,
an edge filter with width of 40 was convolved with the low resolution spectrum in figure 2 .
a strong peak at 6131 is detected in the convolved spectrum , indicating the existence of a lyman limit system at @xmath108 .
no flux is detected blueward of this break .
a lyman limit system is usually detected within 0.1 of the emission redshift in essentially all quasars at @xmath107 ( @xcite , @xcite , @xcite ) .
the most striking feature of the spectrum of sdss 10440125 is the very strong absorption caused by ly@xmath8 forest lines .
however , the flux level in the ly@xmath8 forest region never reaches zero .
it lacks the gunn - peterson ( 1965 ) trough that would exist in the spectrum of a quasar at redshift higher than the re - ionization redshift ( @xcite ) , indicating that the intergalactic medium is already highly ionized at @xmath12 .
we estimate the average continuum decrements as : @xmath109 , where @xmath110 and @xmath111 are the observed and the unabsorbed continuum fluxes of the quasar , and @xmath112 and @xmath113 measure the decrements in the region between rest - frame ly@xmath8 and ly@xmath82 ( @xmath114 ) and between ly@xmath82 and the lyman limit ( @xmath115 ) , respectively ( @xcite ) .
the measurements of @xmath112 and @xmath113 require knowledge of the continuum shape redward of ly@xmath8 .
however , with @xmath112 approaching unity , the effect of different slopes is quite small . assuming a power law continuum @xmath116 with @xmath117 , as indicated in figure 3 , we obtain @xmath9 and @xmath113 = 0.94 . using a slope of 1.0 only changes the @xmath112 and @xmath113 values to 0.92 and 0.96 , respectively .
we therefore adopt @xmath118 and @xmath119 .
these values are in close accordance with those from quasar rd j030117 + 002025 ( @xmath19 , @xcite ) and from distant galaxies in the hubble deep field ( @xcite ) , and are much higher than that of the @xmath120 quasar sdssp j033829.31 + 002156.3 ( @xmath121 , @xcite ) , suggesting that the strong evolution of the strength of the ly@xmath8 forest at @xmath122 , @xmath123 , measured at redshifts below five continues to a redshift of nearly six . assuming this number density evolution , @xcite and @xcite show that at @xmath124 , the expected average @xmath112 ranges from 0.8 to 0.9 .
we further derive an upper limit on the gunn - peterson optical depth following songaila et al .
figure 3 shows the high - resolution echellette spectrum of sdss 10440125 in the ly@xmath8 forest region ( binned to 2 / pixel ) .
the continuum level is approximated by a @xmath125 power law as above .
even the most transparent part of the forest does not return close to the continuum level at this resolution .
it is evident that the ly@xmath8 forest is much stronger in sdss 10440125 than in sdssp j033829.31 + 002156.3 ( figure 2 of @xcite ) , where a fraction of the forest has flux comparable to the extrapolated continuum . in the region between 7926
and 7929 , sdss 10440125 has an optical depth @xmath126 , where the error bar only reflects the statistical noise in the spectrum ( note the noise level indicated in the figure ) .
this value changes to 0.40 and 0.31 for power law slopes of 0.0 and @xmath127 , respectively .
therefore , we adopt a conservative limit of @xmath13 at this redshift of 5.52 . with higher resolution and signal - to - noise ratio , we might be able to select regions even less affected by the ly@xmath8 forest lines .
therefore , it is only an upper limit . for comparison , songaila et al .
derived @xmath128 for @xmath129 with similar resolution .
the gunn - peterson effect analysis above is based on an attempt to measure the amount of flux between ly@xmath8 forest lines ( e.g. , @xcite ) . at redshift
higher than five , even with a moderately high resolution spectrum , these forest lines overlap , making it impossible to find a truly `` line - free '' region . modern hydrodynamic simulations and semi - analytic models show that under the influence of gravity , the intergalactic medium becomes clumpy , and the gunn - peterson optical depth should vary even in the lowest column density regions ( e.g. @xcite , @xcite , @xcite , @xcite ) .
the minimum absorption regions in the forest merely represent regions that are most underdense in this fluctuating gunn - peterson effect .
an accurate measurement of the gunn - peterson effect and the ionizing background from the high resolution spectrum of sdss 10440125 requires detailed comparison with cosmological simulations ; this is beyond the scope of the current paper .
figure 2 also shows the detection of an intervening mgii absorption system .
the mgii doublet @xmath130 is detected at wavelengths 9166.7 and 9190.3 in the high - resolution spectrum ; the redshift of this system is @xmath131 . the rest frame equivalent widths of the doublet lines are 2.43 and 1.90 , respectively .
this system is very similar to the one detected in sdssp j033829.31 + 002156.3 ( @xmath132 , @xcite ) .
it is possible that sdss 10440125 is amplified by lensing from this intervening absorber .
however , we saw in
2 that this object is _ unresolved _ under @xmath133 seeing in the k band .
the total area of sdss imaging data that we have searched for high - redshift quasars thus far is of order 600 deg@xmath16 .
all that satisfy @xmath134 and @xmath135 in this 600 deg@xmath16 region have been observed spectroscopically .
only sdss 10440125 is identified as a high - redshift quasar ; the remaining objects are l and t dwarfs . using the luminosity function and redshift dependence of @xcite ( for the @xmath33 model ) , extrapolating it to higher redshifts and assuming @xmath98
, we predict that in a total area of 600 deg@xmath16 , for @xmath136 ( @xmath1-dropout objects ) , there should be 1.5 quasars with @xmath137 and 1.1 quasars with @xmath138 . for @xmath139 ,
the extrapolation predicts 1.1 and 0.8 quasars for @xmath137 and @xmath140 , respectively .
this assumes that our selection efficiency is 100% .
the schmidt - schneider - gunn luminosity function is derived using objects with @xmath141 and @xmath142 .
although it is difficult to draw any reliable conclusion from the observation of a single high - redshift quasar , the discovery of sdss 10440025 is consistent with the expectations from this rather large extrapolation from lower redshift results .
assuming that the same luminosity function holds at even higher redshift , the sdss will be able to discover one quasar at @xmath143 , @xmath144 in every 1500 deg@xmath16 of the survey .
sdss 1044 - 0125 is a very luminous quasar . assuming that ( 1 ) its bolometric luminosity equals the eddington luminosity , @xmath145 ; ( 2 ) its intrinsic continuum spectrum is the same as the mean spectral template of @xcite , and ( 3 ) neither beaming nor lensing affects the observed flux , we find a black hole mass of @xmath146 in the @xmath29model , or @xmath147 in the @xmath33 model . in either case , the implied black hole mass is quite large , similar to that of the black hole at the center of the nearby giant elliptical galaxy m87 ( harms et al .
1994 , macchetto et al .
if the quasar were radiating below the eddington limit , the inferred black hole mass would be even higher . in the @xcite template , @xmath148 of the bolometric luminosity
is emitted in the observed @xmath149 band .
if this fraction is larger for sdss 1044 - 0125 , the implied black hole mass would be reduced .
note that the universe was less than 1 gyr old at this redshift , while the eddington time scale , the @xmath150-folding time for the growing of a black hole shining at the eddington luminosity , is @xmath151 yr , where @xmath152 is the radiative efficiency for the accretion .
if the black hole started accreting with an initial mass of @xmath153 with 10% efficiency , the seed black hole would have to form and begin accreting at redshift well beyond 10 in order to grow to @xmath154 at @xmath0 ( see also @xcite ) .
forming such a massive black hole in such a short time is a remarkable feat .
the observations of high - redshift quasars can be used to constrain the formation epoch of the first star clusters and the fueling process of early black holes .
how likely is it to find a quasar like sdss 1044 - 0125 in popular cold dark matter cosmological models ?
we use the following simple model to estimate the abundance of high - redshift quasars ( see also @xcite ) : @xcite have found a correlation between central black hole mass and bulge mass , @xmath155 in nearby galaxies . if this correlation holds at high redshift , this would imply a bulge mass of @xmath156 ( @xmath29-model ) or @xmath157 ( @xmath33 ) for the host galaxy of sdss 1044 - 0125 , and a lower limit of @xmath158 ( @xmath29-model ) or @xmath159 ( @xmath33 ) for its dark halo , assuming @xmath160 .
the comoving abundance of dark matter halos at this epoch is very sensitive to this halo mass , and can be estimated by means of the @xcite formalism . in a @xmath29-cdm model , assuming @xmath161 and an untilted primordial power spectrum , for parent halo masses of @xmath162 m@xmath163 , @xmath164 m@xmath163 , and @xmath165 m@xmath163 , we expect 50,000 , 60 , and 4 candidate halos respectively within the survey volume in a redshift window @xmath166
. the duty cycle of quasar activity is poorly known , but is certainly much less than unity ( @xcite ) .
given these uncertainties , this model is not contradictory to our discovery of a single quasar like sdss 1044 - 0125 in the 600 deg@xmath16 survey area .
note that this model does not address the physical process by which the massive black hole formed .
the assumptions we made about @xmath167 and the lifetime of the quasars are completely untested at high redshift . indeed , rix et al .
( 1999 ) argue that the host galaxies of @xmath168 quasars are appreciably less luminous than the universal @xmath169 hypothesis would imply . because of its high luminosity , sdss 1044 - 0125 likely probes the exponential , high mass tail of the underlying dark halo distribution , making predictions of the expected number counts also sensitive to cosmological parameters , especially the normalization of the power spectrum ( i.e. @xmath170 ) . in principle , the sdss survey will be able to probe quasars @xmath171 mag fainter than sdss 1044 - 0125 , and could reveal tens of additional sources at @xmath172 .
the detection of these objects will yield strong constraints on cosmological models for the formation and evolution of quasars at very high redshifts .
the sloan digital sky survey ( sdss ) is a joint project of the university of chicago , fermilab , the institute for advanced study , the japan participation group , the johns hopkins university , the max - planck - institute for astronomy , princeton university , the united states naval observatory , and the university of washington .
apache point observatory , site of the sdss , is operated by the astrophysical research consortium .
funding for the project has been provided by the alfred p. sloan foundation , the sdss member institutions , the national aeronautics and space administration , the national science foundation , the u.s .
department of energy , and monbusho , japan .
the sdss web site is http://www.sdss.org/. this publication makes use of data products from the two micron all sky survey , which is a joint project of the university of massachusetts and the infrared processing and analysis center / california institute of technology , funded by nasa and nsf .
xf and mas acknowledge additional support from research corporation , nsf grant ast96 - 16901 , the princeton university research board , and a porter o. jacobus fellowship .
rhb acknowledges support from the institute of geophysics and planetary physics ( operated under the auspices of the u.s .
department of energy by the university of california lawrence livermore national laboratory under contract no .
w-7405-eng-48 ) .
zh acknowledges support from hubble fellowship grant hf-01119.01 - 99a .
dps acknowledges support from nsf grant ast99 - 00703 .
we thank wolfgang voges , hans - walter rix , david weinberg , and peng oh for helpful comments , and the expert assistance of bob goodrich and terry mcdonald during the keck observations .
esi echellete spectrum over the range 7300 8800 .
the spectrum has been smoothed to a resolution of 2 .
the dotted line shows a @xmath179 continuum , normalized to the region 8650 8730 .
the dashed line indicates the region with minimum absorption ( @xmath180 , see 3.1 ) .
the lower panel shows the noise level . | we present observations of sdssp j104433.04012502.2 , a luminous quasar at @xmath0 discovered from sloan digital sky survey ( sdss ) multicolor imaging data .
this object was selected as an @xmath1-band dropout object , with @xmath2 , @xmath3 .
it has an absolute magnitude @xmath4 ( @xmath5 km s@xmath6 mpc@xmath6 , @xmath7 ) .
the spectrum shows a strong and broad ly@xmath8 emission line , strong ly@xmath8 forest absorption lines with a mean continuum decrement @xmath9 , and a lyman limit system at @xmath10 .
the spectrum also shows strong oi and siiv emission lines similar to those of quasars at @xmath11 , suggesting that these metals were produced at redshift beyond six .
the lack of a gunn - peterson trough in the spectrum indicates that the universe is already highly ionized at @xmath12 .
using a high - resolution spectrum in the ly@xmath8 forest region , we place a conservative upper limit of the optical depth due to the gunn - peterson effect of @xmath13 in regions of minimum absorption . the ly@xmath8 forest absorption in this object is much stronger than that in quasars at @xmath11 .
the object is unresolved in a deep image with excellent seeing , implying that it is unlensed .
the black hole mass of this quasar is @xmath14 if we assume that it is radiating at the eddington luminosity and no lensing amplification , implying that it resides in a very massive dark matter halo .
the discovery of one quasar at @xmath15 in a survey area of 600 deg@xmath16 is consistent with an extrapolation of the observed luminosity function at lower redshift .
the abundance and evolution of such quasars can provide sensitive tests of models of quasar and galaxy formation . |
supersymmetry is up to now one of the most attractive theories beyond the standardmodel ( sm ) .
but from the theoretical point of view there exist many ways for a supersymmetric extension of the sm .
one of the widely used models is the minimal supersymmetric standard model ( mssm ) that has minimal content in the higgs sector .
the masses , mixings and couplings of the four neutralinos are determined by the parameters @xmath5 , @xmath6 , @xmath7 and @xmath8 .
the minimal extension of the mssm by a higgs singlet field @xmath9 is the next - to - minimal supersymmetric standard model ( nmssm ) @xcite .
the neutralino sector now contains five neutralinos being mixtures of the mssm gauginos and higgsinos and an additional singlino @xmath10 . the nmssm neutralino mixing depends on the parameters @xmath5 , @xmath6 , @xmath8 , the singlet vacuum expectation value @xmath11 and the trilinear couplings @xmath12 and @xmath13 @xcite . for large values of @xmath11 the singlino dominated neutralino decouples and the remaining four neutralinos
have the same masses and characters as in the mssm if one identifies @xmath14 @xcite .
the linear collider ( lc ) will first run in the @xmath15-mode .
however , associated neutralino production @xmath16 may not give sufficient evidence for a discrimination between mssm and nmssm , especially if this is the only kinematical accesible neutralino production process in the @xmath15-mode .
in our contribution we want to investigate the possibilities to distinguish between the models with the help of @xmath0 scattering @xmath2 .
we first study the associated production of left / right selectrons and the lightest neutralino ( lsp ) in @xmath0 scattering .
the high energetic photon comes from compton back - scattering of laser photons off one electron beam at a lc .
the production proceeds via s - channel exchange of an electron and t - channel exchange of selectrons .
the formulas for special cases are given in @xcite , the complete analytical expressions for the differential and total cross sections for polarized beams will be given in a forthcoming paper @xcite .
the subsequent decay of the selectrons in electrons and the lsp leads to an @xmath18 and missing energy in the final state .
the total cross section for the production and decay in the @xmath19-cms depends on the polarisation @xmath20 of the electron beam , the helicity @xmath21 of the laser photon and the helicity @xmath22 of the converted electron beam . in fig . 1 we show the contours of constant total cross sections for the process @xmath23 in the @xmath6-@xmath7-plane for the mssm and the @xmath6-@xmath11-plane for the nmssm for @xmath24 gev .
for the calculations we choose the longitudinal polarisation of the electron beam @xmath25 , the helicity of the converted electron beam @xmath26 and the helicity of the laser photon @xmath27 .
this set of polarisations can be achieved at tesla and leads to the most significant contours .
we set @xmath28 , the common scalar mass @xmath29 gev and use the gut relation @xmath30 .
the trilinear couplings are choosen as @xmath31 and @xmath32 ( fig .
1b ) , @xmath31 and @xmath33 ( fig .
1c ) and @xmath34 and @xmath33 ( fig .
1d ) . if in the nmssm the @xmath3 is dominated by the singlino content the production is supressed , because electron and selectron only couple to the gaugino components of the neutralino
. then the total cross section is smaller than 1 fb in some parameter regions ( fig .
1b - d ) and probably lies below the detection limit of a @xmath0 collider .
otherwise one obtains cross sections similar to the mssm . for @xmath35
there also exist a small band with 1 fb @xmath36 10 fb ( fig .
1c / d ) . in this region
the @xmath3 is the mainly singlino - like , but nevertheless it will be very difficult to extract this small signal from the background @xmath37 @xcite .
[ fig1 ] ( 13.9,12.8 ) ( -1.2,1.9 ) ( 3.2,12.6)(a ) ( 3.2,5.9)(c ) ( 10.38,12.6)(b ) ( 10.38,5.9)(d ) ( 6.0,1.9 ) ( -1.2,-4.8 ) ( 6.0,-4.8 )
in this section we consider two cases : ( a ) @xmath3 has singlino character and @xmath38 corresponds to @xmath39 ; ( b ) @xmath40 is singlino dominated and @xmath3 is similar in nmssm and mssm . in table 1
we present four reference scenarios i , ii ( a ) and iii , iv ( b ) with fixed @xmath41 and @xmath42 for comparison . in all scenarios the lightest mssm - like neutralino
is dominated by gaugino components , which couple to electrons and selectrons , while the higgsino components do not . for @xmath43 and @xmath44 ,
the only channel for direct production of neutralinos in @xmath15 annihilation is @xmath45 .
however , if one of the neutralinos is mainly singlino this process will be suppressed .
the cross sections for the process @xmath45 in fig .
2a are calculated with longitudinal polarisations @xmath46 of the electron beam and @xmath47 of the positron beam .
2b / c shows the cross sections for @xmath48 with polarisations @xmath25 , @xmath26 and @xmath27 in the reference scenarios . in all scenarios
we fix the masses @xmath49 gev and @xmath50 gev . .reference scenarios with fixed @xmath41 and @xmath42 ( signs of the mass eigenvalues in brackets ) . in scenario
i and ii the @xmath3 is mainly singlino and in scenario iii and iv @xmath40 is singlino - like . [ cols="^,^,^,^,^",options="header " , ] ... we expect for @xmath51 a production cross section of @xmath52 fb for scenario i and @xmath53 fb for scenario ii at @xmath24 gev ( fig .
the smaller cross section in scenario ii is due to the different relative signs of @xmath41 and @xmath42 , which leads to cancellations in the interference terms @xcite .
now it will be a question of luminosity and background supression , whether the cross sections are measurable .
even if a signal is detected , the decision whether @xmath3 or @xmath40 is the singlino dominated neutralino will be difficult since the cross sections are very similar ( @xmath52 fb in scenario i and @xmath54 fb in scenario iii , see fig .
2a ) . in @xmath0
scattering the process @xmath55 will be strongly supressed ( see sec . 1 ) .
then the first detectable neutralino production process at a high luminosity collider will be @xmath56 .
2c shows total cross sections of @xmath57 fb in scenarios i and ii at @xmath24 gev which further increase for higher energies .
this process , however , will lead to a completely different final state compared to the lsp - selectron production because of the additional decay of the @xmath40 @xcite .
therefore one can decide via detection of the @xmath0 process that the @xmath3 is mainly singlino .
if the @xmath15 cross section is too small for detection , the production and decay of @xmath40 together with @xmath58 in @xmath0 will be the first direct production process of neutralinos that would be measurable at @xmath24 gev with high luminosity .
the non - detection of the @xmath59 production suggests a singlino - like @xmath3 and therefore extended supersymmetry as the underlying theory .
[ fig2 ] ( 13.9,8.8 ) ( 3.2,8.6)(a ) ( 10.38,8.6)(b ) ( 6.8,4.2)(c ) ( -1.2,-2.0 ) ( 6.0,-2.0 ) ( 2.4,-6.4 ) ... we expect for @xmath51 a production cross section @xmath54 fb in scenario iii and @xmath60 fb in scenario iv at @xmath24 gev . if one assumes these cross section to be detectable one again can use @xmath0 scattering to decide which neutralino is singlino dominated . the detection of the process @xmath55 would prove the singlino character of @xmath40 . in scenario iii and
iv one obtains @xmath61 fb for @xmath24 gev ( fig .
if , however , the cross section for associated neutralino production in @xmath15 annihilation can not be observed , the process @xmath55 would be the first process for direct production of neutralinos .
but this process alone does not aalow to discriminate between the mssm and the nmssm .
* for a singlino dominated @xmath3 the associated production of selectrons and the lsp in @xmath0-scattering will be strongly suppressed .
* @xmath62 could be the first process to identify neutralinos and to determine the underlying supersymmetric model . *
if neutralino pair production @xmath63 gives evidence for a neutralino with singlino character , the @xmath0 process can be used to decide , whether @xmath3 or @xmath40 is singlino dominated .
j. ellis , j.f .
gunion , h.e .
haber , l. roszkowski , f. zwirner , phys .
d 39 * ( 1989 ) 844 ; + m. drees , int .
j. of mod .
a 4 * ( 1989 ) 3635 ; + t. elliot , s.f .
king , and p.l .
white , phys . lett .
* b 314 * ( 1993 ) 56 ; phys . rev . *
d 49 * ( 1994 ) 2435 ; + u. ellwanger , m. rausch de traubenberg , and c.a .
savoy , phys . lett . * b 315 * ( 1993 ) 331 ; + b.r .
kim , a. stephan , and s.k .
oh , phys .
* b 336 * ( 1994 ) 200 ; + f. franke and h. fraas , int .
a12 * ( 1997 ) 479 .
f. franke , h. fraas and a. bartl , _ phys .
_ b336 ( 1994 ) 415 ; + u. ellwanger and c. hugonie , _ eur .
phys . j. _ c5 ( 1998 ) 723 ; _ eur .
j. _ c13 ( 2000 ) 681 . j. a. grifols , r. pascual , _ phys .
_ b135 ( 1984 ) 319 .
+ f. cuypers , g. j. van oldenborgh , r. rckl , _ nucl .
_ b383 ( 1992 ) 45 .
+ f. cuypers , g. j. van oldenborgh , r. rckl , mpi - ph/93 - 70 , lmu-93/12 . c. blchinger , h. fraas , in preparation .
s. hesselbach , f. franke , and h. fraas , wue - itp-2000 - 008 , hep - ph/0003272 , contribution to the proceedings of the _
2nd joint ecfa / desy study on physics and detectors for a linear electron - positron collider_. m. baillargeon , g. blanger , f. boudjema , enslapp - a-473/94 . f. franke and h. fraas , z. phys . * c 72 * ( 1996 ) 309 . | we study the associated production of selectrons and neutralinos with subsequent leptonic decay of the selectron in @xmath0 scattering within the framework of mssm and nmssm . due to the weak couplings of singlino dominated neutralinos , associated neutralino production @xmath1 at a linear collider
can be strongly suppressed in some regions of the parameter space .
then the process @xmath2 can give additional informations on the underlying theory and the character of @xmath3 and @xmath4 .
11.30.pb , 12.60.jv , 14.80.ly wue - itp-2000 - 023 + hep - ph / xxxxxxx + , , |
the angular momentum or spin of galaxies is a physical quantity that is far from being fully understood but is of fundamental importance to galaxy formation and cosmological applications . while n - body simulations have shed useful light on spin properties of dark matter halos ( e.g. , * ? ? ?
* ) , it is expected that , given the vastly different scales between the stellar component and dark matter halo component and different physical processes governing stellar , gas and dark matter components , the angular momentum dynamics of galaxies may be quite different and not necessarily inferable from n - body simulations with any reasonable accuracy . we herewith perform a detailed analysis of the dynamics of spin of galaxies in a full cosmological context , utilizing _ ab initio _ laozi cosmological hydrodynamic simulations of the standard cold dark matter model @xcite with an unprecedented galaxy sample size and ultra - high numerical resolution . this paper is the second in the series on the origin of the hubble sequence " .
the reader is referred to @xcite for detailed descriptions of our simulations and validations . briefly
, we perform cosmological simulations with the adaptive mesh refinement hydrocode , enzo @xcite . the periodic box has a size of @xmath1mpc , within which a zoom - in box of a comoving size of @xmath2mpc@xmath3 is emdedded .
the resolution is better than @xmath4pc ( physical ) .
the cosmological parameters are the same as the wmap7-normalized @xcite @xmath5cdm model .
we identify galaxies using the hop algorithm @xcite operating on the stellar particles .
a sample of @xmath6 galaxies with stellar masses greater than @xmath7 are used . for each galaxy at @xmath8 a genealogical line
is constructed from @xmath8 to @xmath9 by connecting galaxy catalogs at a series of redshifts .
galaxy catalogs are constructed from @xmath8 to @xmath10 at a redshift increment of @xmath11 ( corresponding to @xmath12myr at @xmath13 ) and from @xmath10 to @xmath9 at a redshift increment of @xmath14 ( corresponding to @xmath15myr at @xmath16 ) .
the parent of each galaxy is identified with the one at the next higher redshift catalog that has the most overlap in stellar mass .
we compute the specific angular momentum vector @xmath17 for stars of each galaxy within a radius @xmath18 at each output snapshot @xmath19 .
the time derivative of @xmath20 is computed as @xmath21 is somewhat underestimated in cases of rapid changes of angular momentum on time scales shorter than our snapshot intervals .
a similar definition for gas is also used .
we denote @xmath22 as the time required to change the spin vector by @xmath23 degree of arc at each snapshot for each galaxy , defined as @xmath24 where @xmath25 is the unit vector of @xmath17 . for the first time
, we address the evolution of the spin of galaxies statistically in a cosmological setting .
all length units below will be physical .
-0.0 cm -0.5 cm figure [ fig : j3northz ] shows the dot product of the unit vector of the specific angular momentum of the central 3kpc radius stellar region and an arbitrary fixed unit vector as a function of redshift in blue .
it is visible that a significant increase in stellar mass within a short period of time ( i.e. , mergers ) is often accompanied by dramatic changes in angular momentum vectors .
we note that the angular momentum vector of a galaxy over its history displays a substantial amount of change even in quiet " times without major mergers . ensuing analysis provides some physical insight into this .
-0.0 cm -0.5 cm the top panel of figure [ fig : djdt ] shows the pdf of the time derivative of specific angular momentum of the central @xmath26kpc radius stellar regions for galaxies of stellar mass in the range @xmath27 .
the middle panel shows the same as in the top panel , except it is for the central @xmath28kpc radius stellar regions .
we see that the overall rate of change of angular momenta is significantly higher at @xmath29 compared to that at @xmath30 .
the distribution of @xmath31 has an extended tail at the high - end , due to major mergers ; due to our finite time sampling these rates are capped by the frequency of our snapshots .
consistent with the expected decline of major merger rate below @xmath32 , the high @xmath31 tail of the distribution at @xmath30 is significantly less pronounced .
no major difference is seen between @xmath26kpc and @xmath28kpc cases , suggesting that angular momentum changes within the two radii are approximately in tandem and our analysis is robust using @xmath26kpc .
the choice of @xmath33 proper kpc is appropriate by noting that a ( spiral , elliptical ) galaxy of stellar mass @xmath34 is observed to have a size of @xmath35kpc @xcite for low redshift galaxies .
the size roughly scales with the root of the stellar mass and decreases with increasing redshift ( e.g. , * ? ? ?
-0.5 cm -1.0 cm -0.5 cm the bottom panel of figure [ fig : djdt ] shows the pdf for gas in the central @xmath26kpc radius region .
we see that the specific angular momenta of the gas within central @xmath26kpc change at rates @xmath36 times higher than that of stars ( top panel of figure [ fig : djdt ] ) . there is no doubt that gas inflows contribute significantly to the change of the stellar angular momentum in two ways .
first , significant gas inflows at inclined angles to the stellar mid - plane may torque the stars ( and vice versa ) .
second , new gas that reaches there will form new stars that have different angular momentum vector and cause the overall angular momentum to change in both direction and magnitude . at high redshift the orientation of the gas inflows on large scales
are not well correlated with that of the stars or gas that is already there . since
the amount of gas tend to be smaller than that of stars , it is easier to alter the angular momentum of the gas than that of the stars . in the absence of major mergers ,
we expect minor stellar mergers could also alter the angular momentum vector .
figure [ fig : cosdt ] shows the cpdf of the time to change the direction of spin of the central @xmath26kpc radius stellar region by @xmath23 degree of arc , for dependence on mass and redshift ( top panel ) , environment ( middle panel ) and galaxy type ( bottom panel ) .
consistent with figure [ fig : djdt ] we see that the frequency of spin direction change increases with redshift ; the median @xmath37 decreases by @xmath38 from @xmath39 to @xmath40 with the higher mass group corresponding to the high end of the range of change .
the median @xmath37 decreases by @xmath41 from @xmath42 to @xmath43 with the dependence on mass somewhat stronger at low redshift than at high redshift . that less massive galaxies tend to experience more rapid changes of specific angular momenta
is anecdotally apparent in figure [ fig : cosdt ] .
we also find large mis - alignment between inner stellar ( and gas ) regions with outer halos ( not presented here ) , in broad agreement with the conclusions of @xcite .
a dependence on environment is seen in the middle panel , with the median @xmath37 decreasing by a factor of @xmath44 from @xmath45 to @xmath46 ; the environment dependence weakens at higher redshift .
this finding that the spin direction of galaxies changes more frequently in dense environment can be attributed to enhanced local interactions there . in the bottom panel
the dependence on galaxy type gives mixed trends . for blue ( @xmath47 ) galaxies the median @xmath37 decreases steadily from @xmath40 to @xmath39 by a factor of @xmath48 , whereas for red ( @xmath49 ) galaxies the median @xmath37 hardly changes from @xmath40 to @xmath39 .
the median @xmath37 for red galaxies is comparable to that of blue galaxies at @xmath40 ; at lower redshift the median @xmath37 for red galaxies becomes progressively lower compared to that of blue galaxies , mainly due to the latter increasing with decreasing redshift . in @xcite we show that the vast majority of red galaxies do not gain significant stellar mass in the red sequence .
thus we conclude that the rapid change of spin direction for red galaxies are due to torques by nearby galaxies , whereas blue galaxies are subject to all three local interactions - gas accretion , stellar accretion and torques .
it is instructive to put the frequency of spin direction change into some perspective . for a point mass of @xmath50 at a distance @xmath51 , the torque of @xmath50 on the galaxy with a quadrupole moment @xmath52 and angular momentum @xmath53 is @xmath54 ( e.g. , * ? ? ?
* ) , where @xmath55 is the angle between the separation vector and the symmetry axis of the galaxy . expressing @xmath56 in terms of overdensity @xmath57 of the region centered on mass @xmath50 , @xmath58 , where @xmath59 is redshift , @xmath60 the mean mass density at @xmath61 . approximating spirals as flat axisymmetric uniform disks with @xmath62 ( giving the quadrupole moment of @xmath63 ) and full rotation support
this allows to express the torquing time @xmath64 , defined to be the time taken to change the spin direction by @xmath23 degree of arc , @xmath65 giving @xmath66gyr for @xmath67 , @xmath13 and @xmath68 for spiral galaxies . comparing to the median @xmath69gyr seen in figure [ fig : cosdt ] , it is evident that the rapid spin reorientation of galaxies can not possibly be due to tidal torques by large - scale structure .
it is noted that the intrinsic alignment sourced by primordial large - scale gravitational field is inconsistent with the frequent directional change shown in figure [ fig : j3northz ] .
under the ( unproven ) assumption that the large - scale tidal field is the sole alignment agent , any alignment between galaxies on large - scales would result from a balance between the fast reorientation rate due to local processes and slow coherent torques by large - scale structure , which is expressed as the ratio @xmath37 to @xmath64 , denoted as @xmath70 .
if the quadrupole of the galaxy is , in this case , produced by loal interactions , independent of the large - scale tidal field , the alignment in this simplified model would be linear [ instead of quadratic , see @xcite ] to the large - scale gravitational tidal field .
we obtain finally the expression for the mean value of @xmath70 weighted by the distribution of @xmath37 [ @xmath71 , shown in figure [ fig : cosdt ] ] , denoted as @xmath72 : @xmath73 at redshift @xmath59 for galaxies of mass @xmath74 and type @xmath75 ( spiral or elliptical ) .
we approximate elliptical galaxies as oblate axisymmetric spheroids with @xmath76 and @xmath77 , resulting in the quadrupole moment of @xmath78 .
the sizes of galaxies are adopted from observations by @xcite .
the bias factor is from @xcite adjusted to @xmath79 .
the stellar mass to light ratio as a function of absolute magnitude is taken from from @xcite .
we incorporate these into @xmath56 to get @xmath80\ ] ] as well as into @xmath81 in equations ( [ eq : tq ] ) for different galaxy types , where @xmath82 is the linear density growth factor normalized to be unity at @xmath61 , @xmath83 is the bias factor of galaxies of mass @xmath74 ,
-0.5 cm -0.5 cm the results using equation ( [ eq : eta ] ) , in conjunction with equations ( [ eq : tq ] , [ eq : tau ] ) , are shown in figure [ fig : align ] . as in the linear alignment model ( e.g. , * ? ? ?
* ) , the difficulty is to define a demarcating scale between local and linear large - scale structures .
we tentatively have left the scalings to be relative , absorbed into @xmath84 $ ] .
if compelled to give an estimate relevant to weak lensing , one might choose @xmath84 $ ] to be in the range @xmath85 .
in this case , we get a tangential sheer @xmath86 that is @xmath85 times @xmath87 in figure [ fig : align ] , resulting in @xmath86 of @xmath88 for the most massive elliptical galaxies ( i.e. , luminous red galaxies , lrgs , red dots in figure [ fig : align ] ) , which , coincidentally , falls in the range of observed gi ( galaxy - gravitational tidal field ) signal for lrgs ( e.g. , * ? ? ? * ; * ? ? ?
* ; * ? ? ?
the negative sign comes about , because the galaxies , under the torque of a central mass , have a tendency to align their disks in the radial direction that is dynamically stable .
three separate trends with respect to @xmath59 , @xmath74 and @xmath75 are seen : the alignment ( 1 ) decreases with increasing redshift , ( 2 ) decreases with decreasing stellar mass , and ( 3 ) is larger for elliptical galaxies than for spiral galaxies .
the first two trends are accounted for by trends of @xmath37 seen in figure [ fig : cosdt ] .
the last trend requires some discussion .
the bottom panel of figure [ fig : cosdt ] shows that ellipticals have shorter @xmath37 than spirals , due to a large part to their residing in overdense environments and in addition to their having a lower overall specific angular momentum amplitude .
however , also because the specific angular momentum of ellipticals is a factor of @xmath89 lower than that of spiral galaxies , elliptical galaxies are easier to slew .
it is argubly a relatively more straight - forward comparison to observations of ( radial ) alignments of satellite galaxies with respect to the central galaxies of groups and clusters .
but this is in fact complicated by ( at least ) four issues .
first , the observed detection and non - detection of radial alignment of galaxies around groups and clusters of galaxies concern radial ranges that are already mostly in the nonlinear regime ( i.e. , overdensity @xmath90 ) .
second , most of the observed galaxy samples analyzed contain of order 100 - 10000 galaxies , hence statistical uncertainties are in the range of @xmath91 .
third , observed samples likely contain a large number of projected galaxies with physical separations that are much larger than their lateral distance from the cluster / group center ; the degree of projection effects is strongly dependent on the orientation of the line of sight ( e.g. , viewing a cluster along a filament ) and significantly complicates interpretation of results .
fourth , on some very small scales , binary interactions between a satellite and the central galaxy may play the dominant role .
a combination of these factors may explain the current confused state with conflicting observational results ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
nonetheless , we expect that the radial alignment , if exists , is expected to decrease with increasing redshift , perhaps already hinted by some observations ( e.g. , * ? ? ?
* ) , and with decreasing cluster mass at a fixed radius .
the simple model presented has two notable caveats .
first , it assumes that the only alignment mechanism is gravitational torque by some large - scale structure .
so far we have presented only the relative scalings among different galaxies under this assumption , but not the absolute magnitude .
we can not justify this rather critical assumption with confidence at this time .
second , one notes that a significant portion of the galaxy spin direction reorientation is likely due to gas feeding and substructure merging .
thus , it is not unreasonable to expect that the gas feeding and substructure merging have some preferred directions , such as along the filaments and sheets . in this case
, while galaxy spin direction changes frequenctly as shown here , it may do so with some degree of coherence over some scales ( such as the scale of filaments ) , either temporaneously or through long - term memory of large - scale structure ( e.g. , * ? ? ?
if this were true , it then suggests that intrinsic alignments may be a result of balance between high - frequency random re - orientation at short time scales and some sort of large - scale mean " feeding pattern on long time scales .
there is some empirical evidence for galaxies to be aligned with large - scale structures in a sense that is consistent with this feeding " picture ( e.g. , * ? ? ?
* ; * ? ? ?
it should be a priority to understand this issue systematically .
utilizing _ ab initio _ * l*arge - scale * a*daptive - mesh - refinement * o*mniscient * z*oom-*i*n cosmological hydrodynamic simulations ( * laozi simulation * ) of the standard cold dark matter model , we study the evolution of angular momenta of massive ( @xmath0 ) galaxies .
the simulation has an ultra - high resolution of @xmath92pc / h and contains more than 300 galaxies with stellar mass greater than @xmath7 .
we find that spin of the stellar component changes direction frequently , caused by major mergers , minor mergers , significant gas inflows and torques by nearby systems , with a median time in the range @xmath85myr for directional change of spin vector by @xmath23 degree of arc .
the spin of the gas component changes at a factor of @xmath36 higher rates than the stellar component . because the processes that are responsible are mostly in the nonlinear regime .
we do not expect that the findings significantly depend on precise cosmological parameters , the rate of change of spin direction can not be accounted for by large - scale tidal torques , because the latter fall short in rates by @xmath93 orders of magnitude .
in addition , the nature of change of spin direction - apparent random swings - is inconsistent with alignment by linear density field .
a new paradigm emerging with respect to intrinsic alignment of galaxies is that it is determined , primarily , by a balance between slow large - scale coherent torquing ( if it were the sole alignment process ) and fast spin reorientation by local interactions .
this suggests that a significant revision to the large - scale tidal torque based alignment theory is perhaps in order .
a simple analysis presented here indicates that intrinsic alignment of galaxies is dependent on redshift , luminosity , environment and galaxy type .
specifically , it is found that the alignment ( 1 ) decreases with increasing redshift , ( 2 ) decreases with decreasing stellar mass , and ( 3 ) is larger for elliptical galaxies than for spiral galaxies .
while no detailed comparisons are made , the found trends appear to be broadly consistent with and thus provide the physical basis for the observed trends .
what remains open is whether other processes , such as feeding galaxies with gas and stars along filaments or sheets , introduce some coherence of their own kind for spin direction of galaxies along the respective structures .
this will require a separate study in greater detail .
i would like to thank claire lackner for providing the sql based merger tree construction software , and @xcite for providing the very useful analysis and visualization program yt .
computing resources were in part provided by the nasa high - end computing ( hec ) program through the nasa advanced supercomputing ( nas ) division at ames research center .
this work is supported in part by grant nasa nnx11ai23 g .
the simulation data are available from the author upon request . , g. , heckman , t. m. , white , s. d. m. , charlot , s. , tremonti , c. , brinchmann , j. , bruzual , g. , peng , e. w. , seibert , m. , bernardi , m. , blanton , m. , brinkmann , j. , castander , f. , csbai , i. , fukugita , m. , ivezic , z. , munn , j. a. , nichol , r. c. , padmanabhan , n. , thakar , a. r. , weinberg , d. h. , & york , d. 2003 , , 341 , 33 , e. , smith , k. m. , dunkley , j. , bennett , c. l. , gold , b. , hinshaw , g. , jarosik , n. , larson , d. , nolta , m. r. , page , l. , spergel , d. n. , halpern , m. , hill , r. s. , kogut , a. , limon , m. , meyer , s. s. , odegard , n. , tucker , g. s. , weiland , j. l. , wollack , e. , & wright , e. l. 2010 , arxiv e - prints , m. , blanton , m. r. , strauss , m. a. , hoyle , f. , schlegel , d. , scoccimarro , r. , vogeley , m. s. , weinberg , d. h. , zehavi , i. , berlind , a. , budavari , t. , connolly , a. , eisenstein , d. j. , finkbeiner , d. , frieman , j. a. , gunn , j. e. , hamilton , a. j. s. , hui , l. , jain , b. , johnston , d. , kent , s. , lin , h. , nakajima , r. , nichol , r. c. , ostriker , j. p. , pope , a. , scranton , r. , seljak , u. , sheth , r. k. , stebbins , a. , szalay , a. s. , szapudi , i. , verde , l. , xu , y. , annis , j. , bahcall , n. a. , brinkmann , j. , burles , s. , castander , f. j. , csabai , i. , loveday , j. , doi , m. , fukugita , m. , gott , iii , j. r. , hennessy , g. , hogg , d. w. , ivezi , . , knapp , g. r. , lamb , d. q. , lee , b. c. , lupton , r. h. , mckay , t. a. , kunszt , p. , munn , j. a. , oconnell , l. , peoples , j. , pier , j. r. , richmond , m. , rockosi , c. , schneider , d. p. , stoughton , c. , tucker , d. l. , vanden berk , d. e. , yanny , b. , york , d. g. , & sdss collaboration .
2004 , , 606 , 702 , bryan , g. l. , norman , m. l. , oshea , b. w. , abel , t. , wise , j. h. , turk , m. j. , reynolds , d. r. , collins , d. c. , wang , p. , skillman , s. w. , smith , b. , harkness , r. p. , bordner , j. , kim , j .- h . ,
kuhlen , m. , xu , h. , goldbaum , n. , hummels , c. , kritsuk , a. g. , tasker , e. , skory , s. , simpson , c. m. , hahn , o. , oishi , j. s. , so , g. c. , zhao , f. , cen , r. , & li , y. 2013 , arxiv e - prints , i. , frster schreiber , n. m. , rudnick , g. , barden , m. , franx , m. , rix , h. , caldwell , j. a. r. , mcintosh , d. h. , toft , s. , hussler , b. , zirm , a. , van dokkum , p. g. , labb , i. , moorwood , a. , rttgering , h. , van der wel , a. , van der werf , p. , & van starkenburg , l. 2006 , , 650 , 18 | we study the evolution of angular momenta of ( @xmath0 ) galaxies utilizing large - scale ultra - high resolution cosmological hydrodynamic simulations and find that spin of the stellar component changes direction frequently , caused by major mergers , minor mergers , significant gas inflows and torques by nearby systems .
the rate and nature of change of spin direction can not be accounted for by large - scale tidal torques , because the latter fall short in rates by orders of magnitude and because the apparent random swings of the spin direction are inconsistent with alignment by linear density field .
the implications for galaxy formation as well as intrinsic alignment of galaxies are profound . assuming the large - scale tidal field is the sole alignment agent , a new picture emerging is that intrinsic alignment of galaxies would be a balance between slow large - scale coherent torquing and fast spin reorientation by local interactions .
what is still open is whether other processes , such as feeding galaxies with gas and stars along filaments or sheets , introduce coherence for spin directions of galaxies along the respective structures .
= 1 |
the importance of x - ray reflection off the surface of cold matter in the spectra of accreting compact objects has been recognized for a long time .
basko , sunyaev & titarchuk ( 1974 ) discussed the reprocessing of x - rays from an accreting neutron star incident on the surface of the adjacent cooler companion , while guilbert & rees ( 1988 ) considered such reprocessing in clouds " of cold matter in the vicinity of agn .
lightman & white ( 1988 ) studied similar reprocessing on the surface of a geometrically thin , optically thick accretion disk , placing the problem within the `` standard '' paradigm for agn and galactic black hole candidates ( gbhc ) which asserts the presence of such disks illuminated by an overlying hot corona ( e.g. , liang & price 1977 ; galeev , rosner & vaiana 1979 ; haardt , maraschi & ghisellini 1994 ) .
the consequences of x - ray reprocessing in this specific geometry ( arrangement ) have been explored in detail theoretically and observationally .
lightman & white ( 1988 ) and white , lightman & zdziarski ( 1988 ) computed the green s function for the angle - averaged reflected spectrum assuming that the illuminated material is cold , non - ionized , and has a fixed gas density . soon thereafter , a number of authors included the effects of ionization on the structure of the iron lines and the reflected continuum ( e.g. , george & fabian 1991 ; turner et al .
krolik , madau & @xmath7ycki(1994 ) calculated the x - ray reflection and the iron lines from a putative distant obscuring torus , while magdziarz & zdziarski ( 1995 ) calculated angle - dependent reflection component off cold matter ; poutanen , nagendra & svensson ( 1996 ) included polarization in their calculations of the reflected spectra off neutral matter and blackman ( 1999 ) enunciated the influence of the possible concave geometry of the disk on the iron line profile . to date , the most careful ( in terms of the ionization physics ) calculations of the x - ray reflection component and the iron lines from ionized accretion disks in agn are probably those of ross and fabian ( 1993 ) , matt et al .
( 1993 ) , @xmath7yckiet al .
( 1994 ) , and more recently for gbhcs of ross , fabian & brandt ( 1997 ) and ross , fabian & young ( 1999 ) .
all these studies made use of a simplifying assumption , namely that the density in the illuminated gas be constant and equal to the disk mid - plane value .
the justification for this assumption was the fact that in the simplest version of radiation - dominated shakura - sunyaev disks ( shakura & sunyaev 1973 , hereafter ss73 , see their
2a ) , thought to be the case in the majority of the observed sources , the gas density is roughly constant in the vertical direction .
note that the accretion disk is radiation - dominated as long as dimensionless accretion rate , @xmath8 ( @xmath9 is the eddington accretion rate , @xmath10 is the disk bolometric luminosity , @xmath11 is the radiative efficiency of the standard disk in newtonian limit ) , is greater than about @xmath12 .
recently , ross , fabian & young ( 1999 ) suggested that the vertical density structure may follow a gaussian law with the vertical coordinate @xmath13 ( @xmath14 $ ] , where @xmath15 is the scale height , and @xmath16 is the central disk density ) .
they showed that weakness of the observed fe features in spectra of some gbhcs could be interpreted as a result of high degree of ionization of the exposed gas , as opposed to the cold matter intercepting a smaller fraction of the emitted x - rays thereby preserving the paradigm of a cold disk with an overlying hot corona .
however , a self - consistent approach requires that the gas density be determined from the condition of hydrostatic equilibrium _ solved simultaneously _ with ionization , energy balance and radiation transfer equations rather than be assumed . as such ,
the approximation of the constant gas density was made by shakura & sunyaev ( 1973 ) for disks heated by viscous dissipation and for large optical depths only .
therefore , even though the disks we will be considering may be radiation dominated , the constant density assumption may apply only to their deep cooler " layers , where the x - ray heating is negligible . on the contrary , in the upper disk layers ,
this heating exceeds the viscous heating by orders of magnitude ( see [ sect : xstar ] ) , and thus the constant density approximation of ss73 can not be simply extended to these layers .
further , a self - consistent density determination is especially important because of the existence of a thermal instability that allows ( and requires under some conditions ! ) illuminated adjacent gas regions to have vastly different densities and temperatures but , by necessity , similar pressure ( see krolik , mckee & tarter 1981 hereafter kmt ; raymond 1993 ; ko & kallman 1994 ; r@xmath17aska & czerny 1996 , and r@xmath17aska1999 ) . for example , when the gas pressure in the illuminated atmosphere falls below a fraction of the incident x - ray radiation pressure @xmath18 ( @xmath19 is the incident flux ) , the gas temperature may jump from low values ( of the order @xmath20 kelvin for agn ) to the compton temperature , which is typically as large as @xmath1 k. this large jump in temperature makes it impossible to approximate the gas density by either a constant or gaussian profile , as we will see below . in this paper
, we present an x - ray reprocessing calculation that matches or exceeds the most detailed previous works in terms of the radiation transfer and ionization balance calculations , but also relaxes the assumption of the constant gas density .
we adopt a plane - parallel geometry and gravity law appropriate for the standard geometrically thin ss73 disk , and solve for the gas density via hydrostatic pressure equilibrium , taking into account the pressure force due to the x - radiation as well .
the structure of the paper is as follows : in [ sect : s - curve ] we present general considerations on how the thermal instability may influence the x - ray reprocessing and review previously known results . in
[ sect : numerics ] we describe the numerical methods we use to solve the radiation transfer , ionization , energy , and hydrostatic balance equations . readers not interested in these details may skip the latter section and proceed to [ sect : simple ] , where we show several representative tests and compare our results with those obtained with the usual constant density assumption .
radiation - dominated disks are discussed in [ sect : rd ] .
complications arising due to possible x - ray driven gas evaporation in the case of accretion disks with magnetic flares are discussed briefly in [ sect : evaporation ] .
we summarize our results and give our conclusions in
[ sect : summary ] .
the appendix contains notes on the importance of the thermal conduction , two - phase cloudy structure and the intrinsic disk viscous dissipation for the given problem .
in order to simplify our treatment we will assume in what follows that the geometry of the x - ray illuminated gas is a plane parallel one , i.e. , the vertical @xmath13-coordinate is the only dimension relevant to the problem
. the surface of the accretion disk is illuminated by x - rays with a power - law plus exponential roll - over spectrum ( the power law index , @xmath21 , the cutoff energy is fixed at @xmath22 kev ) .
note that due to the fact that most of this radiation is thermalized in the disk and consequently re - radiated , the ionizing spectrum consists , in the zeroth approximation , of the incident x - ray spectrum plus the reprocessed black - body radiation with an equal amount of flux .
the vertical coordinate @xmath13 is counted from the disk mid - plane .
the optical depth in the reflecting material is measured from the top of the illuminated layer , i.e. , @xmath23 , where @xmath24 is the vertical coordinate of the reflector s top , to be found in a full self - consistent calculation .
the existence of multiple temperature solutions for an x - ray illuminated gas at a given pressure was shown by buff & mccray ( 1974 ) and then elaborated by kmt .
it was noted that some of these solutions are unstable .
this instability conforms to the criterion for thermal instability discovered by field ( 1965 ) .
he argued that a physical system is usually in pressure equilibrium with its surroundings .
thus , any perturbations of the temperature @xmath25 and the density @xmath26 of the system should occur at a constant pressure .
the system is unstable when @xmath27 where @xmath28 is the full gas pressure , and the `` cooling function , '' @xmath29 , is the difference between cooling and heating rates per unit volume ( divided by the gas density @xmath26 squared ) . in ionization balance studies , it is convenient to define the `` density ionization parameter ''
@xmath30 , equal to ( e.g. , kmt ) @xmath31 where @xmath32 is the hydrogen density . for situations in which the pressure rather than the gas density is of physical importance
one defines the `` pressure ionization parameter '' as @xmath33 where @xmath34 is the full gas pressure due to neutral atoms , ions and electrons ( we neglect the trapped line radiation pressure in this paper ) .
kmt showed that the instability criterion ( [ field ] ) is equivalent to @xmath35 where the derivative is taken with the condition @xmath36 satisfied , i.e. , when the energy and ionization balances are imposed . in this form
, the instability can be readily seen when one plots temperature @xmath25 versus @xmath37 , since the unstable parts of the curve are the ones that have a negative slope . in figure ( [ fig : scurves ] ) , we show several illustrative ionization equilibrium curves for the case with the ionizing x - ray flux @xmath38 erg sec@xmath39 @xmath40 , photon spectral index @xmath41 and an equal amount of the black - body flux with temperature equal to the effective temperature for the given @xmath19 , i.e. , @xmath42 ev . since these curves produce a shape somewhat similar to the letter `` s '' , they are commonly referred to as `` s - curves '' .
the solid curve shown in figure ( [ fig : scurves ] ) was computed assuming that all resonant lines are optically thin , and we will concentrate on this case for now ( significance of the other curves in fig . [
fig : scurves ] will be explained in [ sect : multiplicity ] ) .
one can see that equilibria in temperature range @xmath43 ev and @xmath44 ev are unstable .
it is also worthwhile pointing out that the instability criterion ( eq . [ [ fcond ] ] ) can be shown to be equivalent to @xmath45 , where @xmath46 is the gas density and @xmath47 is the sound speed ( see nayakshin 1998c , 4.5 . ) an imaginary value of the sound speed in the regions with the negative slope of the ionization balance curve is clearly an `` unusual '' situation , a fact that perhaps helps to clarify the existence of the thermal instability .
a number of authors have investigated the structure of x - ray illuminated atmospheres in the case of a star in a binary system ( e.g. , alme & wilson 1974 ; basko et al . 1974 ; mccray & hatchett 1975 ; basko et al .
1977 ; london , mccray , & auer 1981 ) .
these authors concluded that , as one moves from the star s interior to the atmosphere ( that is from @xmath48 to @xmath49 ) , the gas temperature first follows the s - curve until point ( a ) [ cf .
[ fig : scurves ] ] is reached . for @xmath50 ,
the photo - ionization heating exceeds the cooling and no equilibrium is possible until the gas heats up to a fraction of the compton temperature , @xmath51 kev , where even the high-@xmath52 elements become nearly completely ionized . at this state , the gas cooling and heating
is effected predominantly by compton heating and compton and bremsstrahlung cooling ( kmt ) .
the transition from point ( a ) to point ( b ) is very sharp ] .
in fact , since in the case of a star the escape energy for a proton is around @xmath53 , and the compton temperature @xmath54 few kev @xmath55 , a strong wind results .
mccray & hatchett ( 1975 ) treated the temperature discontinuity as a deflagration wave , in which case the gas pressure is also discontinuous across the ( a - b ) boundary ; it is the momentum flux that is continuous across the transition .
we are primarily interested in the inner part of an accretion disk , where most of the x - rays are produced and where most of the x - ray reprocessing features presumably arise .
because the gas compton temperature , @xmath56 k , is smaller than the virial temperature in the inner disk by several orders of magnitude , no thermally driven wind will escape from the system , and thus there should be a static configuration for the x - ray heated layer in this region ( see begelman , mckee & shields 1983 ) .
thus , the problem of finding the gas density is reduced to finding its hydrostatic equilibrium configuration .
raymond ( 1993 ) and ko & kallman ( 1994 ) have calculated the structure of the accretion disk atmospheres around lmxbs at large radii with special attention to the unknown origin of the uv and soft x - ray emission lines .
the illuminating spectrum was assumed to be thermal bremsstrahlung with @xmath57 k. the work of these authors was pioneering in that they calculated the density of the illuminated gas using hydrostatic balance , thus taking the thermal instability into account consistently .
r@xmath17aska & czerny ( 1996 ) , r@xmath17aska(1999 ) considered the structure of x - ray illuminated disk atmospheres in agn , including the effects of thermal conduction .
the radiation transfer methods used by these last authors did not allow them to study the effects of spectral reprocessing in detail .
our paper is a natural extension of the work of these authors . since we study accretion disks in seyfert 1 galaxies and gbhcs rather than in lmxbs as did raymond ( 1993 ) and ko & kallman ( 1994 ) , we use the ionizing spectrum typical for seyfert 1 galaxies and gbhcs and deal with the innermost part of the accretion disk . compared with work of r@xmath17aska & czerny ( 1996 ) and r@xmath17aska(1999 )
, we solve the radiation transfer `` exactly '' by using the exact frequency dependent cross sections and the variable eddington factor method to take into account anisotropy of the radiation field .
here we describe the part of our code that performs the radiation transfer . the incident x - rays come with a given specific intensity @xmath58 , where @xmath59 is the photon energy and @xmath60 is the cosine of the angle that the given direction makes with the normal to the surface of the disk ( note that @xmath58 is the intensity already integrated over the azimuthal angle , i.e. , it is @xmath61 times the usual radiation intensity ) .
the x - ray intensity is normalized in such a way as to give the total x - ray flux incident on the reflector , @xmath19 ( erg @xmath40 sec@xmath39 ) , integrated over all photon energies and all incident angles : @xmath62 we limit ourselves to treating only @xmath63 upper thomson depths of the illuminated layers , since it is both sufficient and practical ( see [ sect : xstar ] ) .
we break the reflecting layer into @xmath64 ( typically a few hundred ) individual bins .
the binning is the finest around the discontinuity in the gas temperature and density .
the ultimate goal of the radiation transfer is to find the photon intensity , @xmath65 as a function of photon energy @xmath59 , angle ( or @xmath60 ) and position ( equivalently @xmath66 ) everywhere in the slab .
luckily , the ionization balance part of the calculations only depends on the intensity integrated over all angles , @xmath67 : @xmath68 which then allows us to use the variable eddington factors method ( see mihalas 1978 , 6.3 ) .
the variable eddington factors @xmath69 are defined as the ratio of the second to the zeroth moment of the photon intensity : @xmath70 , where @xmath71 is given by an equation analogous to equation ( [ jnu ] ) , but with the additional factor of @xmath72 in the integral .
mihalas ( 1978 ) shows that the transfer equation then reduces to @xmath73 = j - s \ ; , \label{eddap}\ ] ] where @xmath74 is the source function , defined as @xmath75 where @xmath76 and @xmath77 are the scattering and absorption coefficients , respectively : @xmath78 , where @xmath79 is the electron density , @xmath80 is the klein - nishina cross section ; @xmath81 , where @xmath32 is the hydrogen density and @xmath82 is the continuum absorption cross section calculated by xstar ( kallman & mccray 1982 ; kallman & krolik 1986 ) .
further , @xmath83 in equation ( [ eddap ] ) is a differential of the total optical depth , so that it is a function of both photon energy and the location in the slab ( @xmath66 ) and is given by @xmath84 where @xmath85 is the extent of the given bin in the @xmath13-direction .
finally , @xmath86 is the local gas emissivity integrated over all angles , and @xmath87 is the compton scattering operator , discussed below . the compton scattering is treated similarly to the ross & fabian ( 1993 ) treatment of compton down - scattering of incident x - rays : the scattered photons are assumed to be distributed according to a gaussian profile , @xmath88 ( normalized to unity ) , centered at @xmath89 where @xmath90 is the dimensionless electron temperature , @xmath91 is the initial photon energy and @xmath92 .
the energy dispersion of the gaussian is @xmath93 as shown by ross & fabian ( 1993 ) , and as we found in the course of our numerical experimentation , this treatment describes adequately the down - scattering of photons of energy less than @xmath94 kev .
ross & fabian ( 1993 ) used a modified kompaneets equation to describe compton scattering of the low energy photons .
we will see in [ sect : hydro_balance ] that an accurate numerical flux conservation throughout the reflecting layer is a necessary condition for solving the pressure balance satisfactorily .
for best numerical flux conservation , we found that it is preferable to treat all the photons in the same way , avoiding division of the photon intensity on the incident and the ` diffuse ' components .
therefore , we use the same approach ( i.e. , the gaussian emission profile ) for compton scattering of photons of all energies .
this is permissible , since for most of our applications below , the average photon with @xmath95 gains or loses little energy ( @xmath96 ) before it escapes from the layer , and thus compton scattering is essentially monochromatic .
we conducted several tests of comptonization of different photon spectra by slabs of different temperatures ( in the range @xmath97 @xmath98 k , the typical values for our problem ) via both techniques ( i.e. , the gaussian emission profile and the kompaneets equation ) and found no noticeable differences in the results for the slabs of thomson depth @xmath99 few ( but see the end of [ sect : summary ] ) .
accordingly , the result of the compton scattering operator acting on @xmath100 in our approach is @xmath101 where @xmath102 is the intensity of once scattered photons found by convolving the gaussian re - distribution profile @xmath103 with the unscattered intensity @xmath100 : @xmath104 the radiation transfer equation ( [ eddap ] ) is supplemented by boundary conditions . on the top of the illuminated layer ( i.e. , for @xmath105 ) , we require that the down propagating flux is equal to the incident x - ray flux : @xmath106 = f_x(e)\ ; , \label{bct}\ ] ] where @xmath107 is the ratio of the first to the zeroth moment of the intensity at @xmath105 ( see mihalas 1978 , 6.3 , especially text after equation ( 6 - 42 ) ) .
we assume that all of the incident x - ray flux is reprocessed into the thermal disk flux deep inside the disk ( e.g. , sincell & krolik 1997 ) , and that there is also an intrinsic disk flux @xmath108 whose magnitude is simply given by the formulae of ss73 .
we found that the most physical and numerically accurate boundary condition is a mirror boundary condition for @xmath109 , similar to that used by alme & wilson ( 1974 ) , with the appropriate correction due to the additional disk flux @xmath108 : @xmath110 = f_d(e)\;. \label{bct1}\ ] ] if @xmath111 , then equation [ bct1 ] leads to the exact mirror boundary condition , i.e. , every photon is reflected without any change in its energy .
this is somewhat unphysical because in reality , due to photo - absorption , x - ray albedo is less than unity , but since only a tiny fraction of x - rays will reach the bottom of the reprocessing layer , we are committing a small error .
the variable eddington factors are obtained in the usual way ( mihalas 1978 , 6.3 ) .
one defines the function @xmath112 as @xmath113/2 $ ] for @xmath114 .
the angle dependent equation of the radiation transfer can then be written as @xmath115 and it can be solved _
exactly _ once @xmath100 and the source function , @xmath74 , are known .
the solution of this equation allows us to refine our initial estimate for the variable eddington factors @xmath69 and the flux factors @xmath116 .
these new functions are then employed in equation ( [ eddap ] ) to obtain the next iteration of @xmath100 .
we repeat this cycle without calling xstar ( i.e. , with fixed opacities and emissivities obtained from the last xstar call ) typically 30 times such that the eddington factors converge to a good degree .
our treatment of the thermal and ionization balance is closest to the work of @xmath7yckiet al .
as mentioned earlier , we use the photoionization code xstar ( kallman & mccray 1982 ; kallman & krolik 1986 ) .
we start from the first zone situated on the top of the reflecting gas and work our way to the bottom of the layer .
we provide xstar with the ionizing intensity , @xmath100 , for each zone computed in the previous iterations or given by the initial value for the first iteration . the resonance line optical depth for photons emitted in the first layer
is assumed to be zero for each line included in xstar .
the latter then provides us with line opacities for this zone , which we use to compute the line optical depths for the line photons emitted in the next zone .
we repeat this process for every spatial bin ; therefore , the optical depth of a particular resonant line is a sum of the appropriate line depths for all the previous zones above the given one .
the treatment of the coldest part of the reflecting layer deserves special attention . in terms of the number of the included elements , ionization states and excited states
xstar is one of the most extensive codes existing at the present time .
however , the code does not include stimulated emission or de - excitation processes of certain excited states , so it may incorrectly estimate the cooling rate under some conditions . as in raymond ( 1993 ) and @xmath7yckiet al .
( 1994 ) , we set the gas temperature to the effective one ( @xmath117 ) , if the value computed by xstar is lower than @xmath117 .
clearly , this approximation is not very accurate for photon energy @xmath118 few @xmath119 , and may not give correct opacities and emissivities in that energy range .
the focus of this paper , however , is on the x - ray reflection feature and the iron line , i.e. the photon energy range @xmath120 kev , so that the approximation described above should suffice .
further , in most previous studies , one usually sets a constant black - body or modified black - body continuum to propagate in the upward direction ( e.g. , ross & fabian 1993 ; ko & kallman 1994 ; @xmath7yckiet al .
since we intend to include the radiation pressure in the hydrostatic balance equation , this approach is not suitable for us .
we need to propagate the quasi - thermal component through the layer just as we do the x - rays .
the difficulty with this is that we previously introduced the minimum temperature @xmath121 , which forbids xstar to solve the energy balance equation if it leads to @xmath122 .
therefore , the opacities and emissivities obtained from xstar in this way will not match each other ( i.e. , total energy absorbed @xmath123 total energy emitted ) . to cope with this problem ,
we redefine the emissivities obtained from xstar in this regime in the following manner : when @xmath121 , we count the line emission only if the line energy @xmath124 ev , otherwise we set it to zero ( because we are mostly interested in the x - ray emission ) .
the continuum emission is assumed to be given by @xmath125 , where @xmath126 is the planck function at @xmath127 .
this choice of @xmath86 is appropriate since it is exactly the kirchoff - planck law and is often adopted as an expression for the rate of thermal emission . the quasi - thermal emission spectrum @xmath86 is renormalized in such a way as to satisfy the energy equilibrium equation .
finally , one practically important question is how deep in the illuminating layer should we carry our calculations ( i.e. , how large should @xmath128 be ? ) . in terms of the energy balance for the illuminated layer ,
it is sensible to extend the computational domain down to the thomson `` viscous heating depth '' , @xmath129 , such that for @xmath130 the intrinsic disk viscous heating wins over the x - ray heating . to estimate value of @xmath129
, we proceed in the following manner .
the intrinsic disk heating per hydrogen atom , in the framework of ss73 model , is @xmath131 , where @xmath132 is the mid - plane hydrogen density .
further , we follow sincell & krolik ( 1997 ) , who derived a useful approximate expression for the x - ray heating rate ( see their equation 32 ) .
re - writing it in terms of thomson depth in the limit @xmath133 , @xmath134\;. \label{qx}\ ] ] hence , the ratio of the x - ray heating rate to viscous dissipation is @xmath135 where @xmath136 is the total thomson depth of the accretion disk . clearly , the x - ray heating exceeds the viscous heating by orders of magnitude for @xmath137 , unless @xmath138 , a situation of no particular interest to us .
accordingly , we neglect the viscous heating in the illuminated layer ( although see the appendix ) .
further , it is computationally impractical to extend computational domain to @xmath139 , since the computing resources are limited .
fortunately , very few x - rays penetrate to thomson depth greater than @xmath140 few .
therefore , we choose to treat with the full formalism developed in this paper only the upper 4 thomson depths .
in addition , in order to reproduce the thermal component in the reflected spectra , one ought to have some material at that temperature ( since our mirror boundary condition does not by itself create thermal flux in the limit @xmath141 ) . in the realistic situation , this material lies below @xmath142 , according to the foregoing discussion .
since we limit ourselves to @xmath143 , we instead require that the gas temperature always be the effective temperature for @xmath144 . via numerical experimentation
, we found that the thomson depth of @xmath145 is indeed adequate to compute the reprocessing features , and that the region @xmath146 is sufficiently optically thick to produce the thermalized part of the reflected spectrum even if @xmath147 .
these restriction on @xmath66 does not affect our results at all . as we found ( see fig .
[ fig : temper1 ] ) our highest ionization case becomes cold " below @xmath148 . as apparent in fig .
( [ fig : spectra1 ] ) in this specific case the associated reprocessing features are extremely weak .
therefore , in a situation in which the ionized layer extends to @xmath149 such features would be completely unobservable . currently , one of the most widely considered models of the x - ray emission from accretion disks is that where x - rays come from localized active regions , thought to be magnetic flares ( e.g. , galeev , rosner & vaiana 1979 ; haardt , maraschi & ghisellini 1994 ; svensson 1996 ; nayakshin 1998a , b , c ) . in the framework of that model ,
the x - ray flux from the active region is very much larger than the radiation flux generated within the disk .
nayakshin & melia ( 1997a ) , nayakshin ( 1998b ) and nayakshin & dove ( 1999 ) found that the radiation pressure from the x - rays of magnetic flares may be larger than the unperturbed gas pressure of the accretion disk atmosphere .
thus , we ought to incorporate the effects of the incident radiation pressure in our pressure balance calculations for generality . the radiation force per hydrogen atom inside the reflecting layer , @xmath150 , is equal to @xmath151 where we introduced @xmath152 , which has dimensions of a cross section ( @xmath153 here is the total cross section per hydrogen atom ) . note that this quantity can be positive as well as negative ; a positive @xmath154 corresponds to radiation force pointing upward , which would happen when the cross section for the interaction of matter with the reprocessed emission is larger than that with the x - rays .
the equation for the pressure balance now reads @xmath155 it is convenient to re - write this equation in terms of dimensionless quantities that will figure prominently in the ionization and radiation transfer calculations .
these are : the gas pressure normalized by the incident x - ray radiation pressure , @xmath156 ; the differential thomson depth , @xmath157 ; and the compactness parameter @xmath158 defined as @xmath159 we will further introduce a dimensionless `` gravity parameter '' , @xmath3 , for brevity of notations : @xmath160 where @xmath161 is defined by @xmath162 and @xmath163 is the scale height of the ss73 accretion disk .
the parameter @xmath3 characterizes the strength of the gravity term relative to that of the x - radiation pressure in equation ( [ heq1 ] ) .
the hydrostatic balance equation is given by @xmath164\ ; , \label{heq2}\ ] ] where @xmath165 .
the solution of the hydrostatic balance equation ( [ heq2 ] ) is subject to boundary conditions .
the boundary condition on the top of the illuminated layer is @xmath166 , obviously ( although note that if a hot corona is placed on the top of the illuminated layer , then @xmath167
see r@xmath17askaet al .
practically , we set @xmath168 to a small fraction of the incident x - radiation ram pressure , i.e. , @xmath169 .
the boundary condition on the bottom of the illuminated layer requires that the gas pressure be continuous across the interface between the illuminated layer and the accretion disk .
once the pressure profile in the disk is known , it is a relatively simple matter to iterate on the _ a priori _ unknown location of the top of the illuminated layer @xmath24 in order to adjust the gas pressure on the bottom to match the disk pressure .
however , one inherent problem here is that the vertical structure of the accretion disks is not necessarily well known ( see , e.g. , de kool & wickramasinghe 1999 ) . for the gas - dominated accretion disks
, we will simply assume that the density profile follows exponential law by our choice .
] : @xmath170\ ; , \label{rhogd}\ ] ] where @xmath171 is the `` average '' scale height , which we approximate as @xmath172^{1/2}$ ] , where @xmath173 is the boltzmann constant and @xmath174 is the disk mid - plane temperature . in principle
, one could resolve the vertical structure of the disk to a `` better '' precision by integrating the hydrostatic balance equation from the disk mid - plane up .
however , the distribution of the disk viscous dissipation with height is not known ( ss73 simply assumed it is proportional to the gas density ) , precluding the accurate determination of the temperature profile @xmath175 of the disk . fortunately , as we will see later , this does not bring any serious complications except for a slight uncertainty in the exact value of @xmath176 .
we are less lucky in the radiation - dominated ( rd ) case , since there the boundary conditions are of a greater importance . to see this ,
consider the case in which the illuminated layer is completely ionized , so that the only source of opacity is the thomson one . in this case , @xmath177 , and the second term on the right in equations ( [ heq1 ] ) and ( [ heq2 ] ) is zero .
further , note that without the second term , equation ( [ heq1 ] ) is exactly the equation for hydrostatic balance inside the disk . in the rd case ,
the gas pressure is negligible everywhere except for very near the edge of the disk . in other words , inside the disk
, the gravity term is canceled by the intrinsic radiation pressure term ( the last term in eq .
[ heq1 ] on the right ) .
as we found , the illuminated layer does not extend very high in the rd case , e.g. , @xmath178 , so that the first term on the right in eq .
( [ heq1 ] ) is again almost completely canceled by the last term . in other words , the net force is the difference of two very large terms , and hence it is rather sensitive to the location of the illuminated layer lower boundary @xmath176 .
another serious problem is that the gas density profile in an rd disk is uncertain to a much greater degree than in the case of a gas - dominated ( gd ) disk .
as shown in ss73 , if one completely neglects the gas pressure , then the density is constant with height all the way to @xmath179 in rd disks . in order to significantly improve the ss73 approximation to the vertical disk structure
, one needs to solve the energy balance equation for the accretion disk _ simultaneously _ with the pressure balance .
this is a more difficult problem ( and more model dependent ) than the one we are attempting to solve in this paper ( e.g. , de kool & wickramasinghe 1999 ) , and for this reason it is clearly outside of the scope of our present work .
therefore , at present , we make the _ total _ pressure continuous across the lower boundary and use the approximation of ss73 for the vertical pressure structure .
since we require the gas temperature to be equal to the disk effective temperature and since the radiation field incident on the lower boundary from the illuminated layer is quite close to the planckian function , the radiation pressure is actually continuous through the boundary by our choice ( see discussion below eqs .
[ bct ] and [ bct1 ] ) .
thus , if we match the total pressure on both sides of the boundary to within a very small difference , the gas pressure and density will also be continuous . in order to achieve static solutions for all the quantities in all vertical zones
, we make the following steps : \1 ) first , we guess the location of the upper boundary of the illuminated gas @xmath24 , and we also assume an initial intensity @xmath180 in all the spatial bins , usually simply equal to the sum of the incident x - ray intensity and the black body spectrum , and use it as an input to xstar in order to obtain the opacities .
the initial value of the eddington variable factors is assumed to be the isotropic value , i.e. , @xmath181 .
then , starting from the top of the illuminated layer , we solve for the gas pressure in the next zone using eq .
( [ heq2 ] ) .
given that value of @xmath34 , we call xstar with the initial illuminating spectrum .
xstar provides on output the gas temperature , density , opacities , line and continuum emissivities for this zone .
we repeat this step for the next zone ( i.e. , again solve for its pressure and then call xstar ) , until we reach the bottom of the transition layer .
\2 ) having found the gas opacities and emissivities , we proceed to compute the radiation transfer as described in
[ sect : rad_transfer ] . at this step ,
we fix the opacities and emissivities at the value given by the previous iteration of xstar , and make 20 - 25 iterations of the radiation transfer calculations ( made by an implicit differencing scheme ) until the reflected spectrum converges .
the variable eddington factors are also continuously updated during this step .
\3 ) at the same time , we check the consistency of the lower pressure boundary condition . in particular , we compare the pressure at the base of the illuminated layer ( i.e. , at @xmath182 ) with the ss73 pressure at the same height . if the latter is larger than the former , we increase the value of @xmath24 from the previous iteration ; if not , we decrease it .
\4 ) now that we have the refined value of @xmath24 , and the new ionizing spectrum in each bin , we repeat step ( 1 ) using the new values for @xmath24 , the ionizing spectrum and the variable eddington factors . we continue to repeat these steps until a convergence is reached .
we found that the best indicator of convergence is the matching of the boundary pressure at the base of the layer , @xmath183 to the ss73 pressure at that height @xmath184 .
we typically require that @xmath185 .
we also check that the temperature structure of the illuminated layer stops evolving to the same level of precision .
typically , we need to repeat the whole cycle 20 or more times , which translates into @xmath118 a day of cpu of a 500 mhz pc for about 250 spatial zones and 400 photon energy points .
figure ( [ fig : scurves ] ) shows that due to the thermal instability , there are two , and sometimes even three stable solutions of the thermal and ionization balance equations for a given gas pressure and local ionizing spectrum when the gas pressure is in the range @xmath186 ( equivalently , @xmath187 ) .
thus , one _ has to choose _ a solution out of the several possible based on additional considerations .
the approach that would leave no ambiguities would be to supplement our ionization and radiation transfer methods with dynamical equations for the gas motions in the illuminated layer ( e.g. , alme & wilson 1974 ) .
such an approach would delineate the different solutions automatically , since then one would follow the evolution of the gas density explicitly . for a given density ( unlike given pressure ) , there is always a thermally stable , unique solution of the ionization and thermal balance equations ( this can be seen from an always positive slope of the @xmath25 versus @xmath30 curve ; see , e.g. , nayakshin 1998c , fig .
in addition , inclusion of thermal conduction ( e.g. , r@xmath17aska1999 ) can also forbid some of the otherwise stable solutions ( see below ) .
unfortunately , the numerical complexity of such an approach is prohibitive .
the problem with adding the heat conduction to our code is that in our iterative process , we move from a known temperature in zone @xmath188 to solve for temperature in zone @xmath189 , whereas thermal conduction requires knowledge of temperature in zone @xmath190 at this step .
writing a time - dependent code that would self - consistently treat the line cooling is also challenging and beyond the scope of the present work . given this , we should be able to find a way to single out a solution with some desirable property ( e.g. , within a given temperature range ) out of several possible solutions via our numerical methods .
we discovered that the easiest way to accomplish this is to give xstar an initial temperature guess , @xmath191 that is somewhat close to the temperature of the solution that we want to pick .
this is due to the fact that xstar defines the temperature range where it will seek a solution in some interval around the initial temperature @xmath191 .
if there indeed exists a solution that is close to @xmath191 , then xstar converges to that solution ; if that solution does not exist , then xstar expands the interval until it finds a solution . to avoid complications due to line cooling ,
we first demonstrate this method of separating the solutions with the example of a calculation in which all the resonant lines were assumed to be optically thin , i.e. , @xmath192 for all lines in all spatial zones .
figure ( [ fig : thin ] ) shows the gas temperature as a function of the thomson optical depth for two tests . in both tests ,
the x - ray flux with @xmath193 erg s@xmath39 @xmath40 is incident in the direction normal to the reflecting layer .
other parameters have the following values : @xmath194 , @xmath195 , and @xmath196 ( given that , one can compute the disk vertical height scale @xmath163 ; for simplicity , we assume that the bottom of the reflecting layer is located at @xmath197 for the tests in this section ; this does not affect any of our conclusions here ) .
the solid curve was computed for the case when an initial temperature guess for each zone is equal to @xmath198 kev .
this method ensures that whenever there are multiple solutions , the hottest one is chosen . in terms of the optically thin s - curve shown in figure ( [ fig : scurves ] ) , this simulation yields the transitions from the hot compton - heated branch to the medium stable branch via line ( c - d ) , and then via line ( e - f ) .
the dotted curve shows results of a test when an initial gas temperature is close to @xmath199 .
as we found , this method always leads to a solution with the smallest possible temperature , which corresponds to the discontinuity occurring via line ( a - b ) in figure ( [ fig : scurves ] ) . in figure
( [ fig : thick ] ) we show three tests for the same parameter values , but with the resonance line optical depths self - consistently calculated as described in [ sect : xstar ] .
the long - dashed curve shows the results of a calculation with the initial temperature for each bin @xmath200 . as with the dotted curve of fig .
( [ fig : thin ] ) , the cold solution is recovered rather quickly at @xmath201 , as soon as it becomes available .
contrary to the optically thin case shown in fig .
( [ fig : thin ] ) , however , the next zone jumps back to the hot solution .
the reason for this is line cooling . for a given bin ,
the line optical depths are calculated by summing over the line depths of all the previous bins .
since the material is highly ionized for all zones preceding the one with @xmath200 , there is little line optical depth for this zone , and thus we can use the s - curve generated for optically thin lines ( solid curve in fig . [
fig : scurves ] ) .
however , for certain lines the optical depth of the first cold zone turns out to be significant , and therefore cooling for all the material with @xmath202 can not be assumed to be ( line- ) optically thin anymore .
we then need to refer to the other two curves in fig .
( [ fig : scurves ] ) that were calculated for a non - zero line optical depth @xmath203 ( where it was assumed for simplicity that all the lines have a given value of @xmath203 ) .
in particular , the longed - dashed curve in figure ( [ fig : scurves ] ) , shows that there is no cold solution with @xmath204 , where @xmath205 is the pressure ionization parameter at point ( b ) .
the only acceptable solution is the hot solution , which then explains why the temperature in the next zone jumps back to the compton equilibrium .
note that a finer ( than the one we used ) zoning around this point will not remove this behavior , since this behavior is physical rather than numerical .
the two other dips in temperature around @xmath206 and @xmath207 in the dashed curve in fig .
( [ fig : thick ] ) are caused by similar effects due to different lines .
since the sharp dips in the temperature will cause a strong conductive heating , the cold temperature guess method seems to be inappropriate .
further , as discussed by london et al .
( 1981 ) , the effects of the line cooling may lead to a non - steady behavior in the illuminated layer .
indeed , the layer at @xmath201 is very much denser than the layer immediately below it , and thus it is convectively unstable .
the dotted curve in figure ( [ fig : thick ] ) was computed assuming always high initial temperature ( @xmath208 kev ) .
this curve produces one sharp dip in temperature at @xmath209 . at this point ,
only the cold solution exists . in the next spatial zone , due to the larger line depth
, there appears a solution with the intermediate temperature , @xmath210 ev , which is found by xstar since the initial guess for @xmath25 is high .
however , there also exists a colder solution , which is seen in the solid curve presented in the same figure .
the latter curve was computed by providing xstar with the initial temperature equal to that of the previous zone multiplied by @xmath211 . if the conductive heating was taken into account , it would force the gas in the zone with the large jump in temperature to cool and so it would lead to a solution close to the solid line .
we show in the appendix that the conductive heating is always small except for regions that are close to the temperature discontinuities .
the latter regions are always optically thin ( @xmath212 at most , see appendix ) , and hence they are unimportant from the point of view of the radiation transfer and spectral reprocessing . out of these three methods of picking a solution , the one based on the temperature of the previous zone is the most physical in the sense that it has no sharp dips or spikes in temperature except for the bump with the magnitude of about 2 in the region around @xmath213 ( similar features were earlier seen in the work of ko & kallman 1994 ) .
therefore , we adopt this method in the rest of the paper .
a systematic investigation of the reflected spectra as a function of the accretion rate , @xmath214 , and the magnitude of the x - ray flux , @xmath19 , is our eventual goal . solving this problem
consists of two steps : ( 1 ) resolving the density and temperature profile of the illuminated layer ; ( 2 ) determination of how this profile influences the radiation transfer in the illuminated layer , i.e. , the reflected spectra . step ( 1 ) is not necessarily difficult , but it depends quite sensitively on the assumptions about the geometry of the x - ray emitting source ( or sources ) .
for example , if x - rays are produced in magnetic flares , then , because the geometry of the problem is no longer plane - parallel , there is a possibility that a local wind is induced ( see [ sect : evaporation ] on this ) .
we clearly can not treat this situation within our current static approach .
another possibility is that the accretion disk is not described by the standard model , such that @xmath215 is different from the ss73 value , which then changes the gravity parameter @xmath3 . in other words
, we have a large parameter space to investigate , and such an investigation clearly goes beyond the scope of this paper . at the same time , for a fixed @xmath0 , step ( 2 ) depends , to a large degree , on two parameters only .
one of these is the thomson depth of the compton - heated layer , @xmath216 , and the other is the ratio of the illuminating flux to the disk intrinsic flux , @xmath217 , since this ratio determines the actual value of the compton temperature @xmath218 .
therefore , as a first attempt towards understanding of the effects of the thermal instability on the reflected spectrum , we set up simple tests that allow us to cover a range of values of @xmath216 and @xmath218 by varying one parameter at a time , thus isolating dependence of the reprocessed spectra on that particular parameter .
we assume the following initial setup .
we fix the accretion rate at a value of @xmath219 ( the disk is gas - dominated for such a small accretion rate ) , the incident x - rays to be isotropic for simplicity , the coronal dissipation parameter @xmath220 ( no dissipation in the corona ; see svensson & zdziarski 1984 for a definition ) , the black hole mass to @xmath221 , and the incident x - ray flux ( projected on the disk surface ) to @xmath193 ergs s@xmath39 @xmath40 . for these values of parameters , @xmath222 , and the gravity parameter @xmath3 equal to @xmath223 . to study the effects of the thermal instability ,
we artificially vary the parameter @xmath3 from this `` true '' value ( other parameters of the problem will later be varied as well ) .
the rest of our methods , i.e. , the radiation transfer , the boundary conditions between the disk and the illuminated layer are unchanged , that is the same as we will later use in fully self - consistent calculations ( [ sect : rd ] ) .
it is beneficial to start by comparing the results of our self - consistent calculations with those obtained with the fixed density approach , since the latter dominated the literature on the topic during the past decade ( e.g. , ross and fabian 1993 ; @xmath7yckiet al .
1994 ; ross et al . 1997 , 1999 ) .
the most important parameter in photo - ionization calculations when the gas density is constant is the density ionization parameter @xmath30 ( eq .
( [ xid ] ) . in the case of our calculations that include the hydrostatic balance , the ionization structure of the illuminated layer depends most sensitively on the `` gravity parameter '' @xmath3 .
the upper panel of figure ( [ fig : temper1 ] ) shows temperature of the illuminated layer as a function of the thomson depth into the layer for several values of the gravity parameter @xmath3 , whereas the lower panel of the same figure shows the gas temperature for different values of the density ionization parameter @xmath30 ( both the fixed density cases and the self - consistent cases are computed with our code under equal conditions and parameter values except for the gas density structure ) . in short , the two sets of temperature curves have almost nothing in common .
the self - consistent solution avoids the regions in temperature that correspond to a negative slope in the `` s - curve '' ( cf .
[ fig : scurves ] ) , i.e. , the unstable regions , whereas they are unphysically present in the fixed density cases ( fig .
[ fig : temper1 ] ) . to see that the fixed density solutions are indeed unphysical
, notice that since the gas density is constant , the gas pressure is directly proportional to its temperature .
hence , on the top of the illuminated atmosphere , instead of decreasing to zero , the gas pressure exceeds that on the bottom of the layer by a factor as large as a few hundred .
the constant density calculations are out of the pressure balance , and therefore will not exist in reality .
the temperatures that are allowed in the self - consistent solutions are those with the positive slope in the `` s - curve '' ; the amount of thomson depth they occupy depends on the pressure balance .
going from the low illumination ( high @xmath3 ) cases to the high illumination limit ( @xmath224 ) , the amount of the hot material increases , as it should , but the temperature of the compton - heated layer does not change considerably , except for the case when the hot layer becomes thomson thick .
if one assumes that the hot material is ionized enough to neglect line and recombination cooling , then it is possible to analytically find the temperature , @xmath225 , below which the compton heated solution ceases to exists ( see ivb in kmt ) .
the value of @xmath225 is @xmath226 , i.e. , a third of the compton temperature at the very top of the illuminated layer .
the upper panel of figure ( [ fig : temper1 ] ) shows that this transition temperature is indeed close to a third of the maximum temperature if @xmath227 . in the opposite limit
, the compton - heated layer is optically thick and thus the ionizing spectrum is actually different at the surface and at the transition ; the simple analytical theory would need to take this fact into account in order to be more accurate .
in contrast , the fixed density calculations incorrectly predict that not only the thomson depth of the hot layer , but its temperature , too , changes . this fact has obvious consequences for the reflection component and the iron lines , as we will see below .
figure ( [ fig : structure ] ) shows the gas temperature , the `` radiation pressure force multiplier '' @xmath228 , the gas pressure and the height above the bottom of the illuminated layer for the self - consistent test with @xmath229 .
( [ fig : structure]d ) , in particular , demonstrates that the density profile can not be assumed to be a gaussian , because , in a rough manner of speaking , there are two height scales
one for the cold part of the illuminated layer , and another for the compton - heated part .
comparing the temperature profiles shown in the upper panel of fig .
( [ fig : temper1 ] ) with results of ko & kallman ( 1994 ) , r@xmath17askaand czerny ( 1996 ) , r@xmath17aska(1999 ) , one finds that the intermediate stable solution occupies a smaller thomson depth relative to the compton - heated solution than found by these authors .
this small thomson depth is also somewhat unexpected because the part of the s - curve between points ( e ) and ( d ) in fig .
( [ fig : scurves ] ) is not `` small '' , i.e. , it corresponds to a change in the gas pressure by a factor of @xmath230 few .
panel ( b ) of figure ( [ fig : structure ] ) shows that @xmath231 everywhere except for the region right below the interface between the hot and the cold solutions , where the x - ray radiation pressure force has a sharp peak .
the direction of the force is always down , i.e. , in the same direction as the disk gravity , because the soft x - ray opacity dominates over the opacity to the quasi - thermal disk spectrum .
panel ( c ) of figure ( [ fig : structure ] ) shows that due to this x - ray pressure , the gas pressure increases almost discontinuously once the temperature falls to the cold ( or the intermediate ) stable branch values .
thus , the intermediate stable branch of the s - curve can be passed very quickly and even `` skipped '' because of the additional radiation pressure .
also note that the radiation pressure force due to the incident x - rays minus that due to the reprocessed flux is small in the integrated sense .
this allows us to approximately neglect this term in the hydrostatic balance equation ( [ heq2 ] ) .
if the transition from the hot to the cold solution occurs at @xmath232 few , then equation ( [ heq2 ] ) tells us that the thomson depth of the hot layer is roughly @xmath233 in accordance with this simple estimate , fig .
( [ fig : temper1 ] ) shows that larger values of @xmath3 result in lower values of @xmath216 , although the dependence of gravity law on the height @xmath13 makes this behavior more complex than equation ( [ tauh ] ) predicts .
a much better analytical theory describing the temperature structure arising from the self - consistent calculations can be obtained by using the analytical theory of kmt and the equation for the hydrostatic balance ; we will present this theory in a future paper .
we will now discuss the reprocessed spectra that correspond to the temperature profiles shown in fig .
( [ fig : temper1 ] ) . figure ( [ fig : spectra1 ] ) shows the reflected spectra for both the constant density cases and our self - consistent calculations in the important medium and hard x - ray energy range .
these spectra are shifted with respect to each other to allow easy visibility of the line . also , in order to expose the true strength of the line and the reflected continuum , we re - plotted several of these curves with their correct normalization in fig .
( [ fig : spectra2 ] ) .
finally , for completeness , fig .
( [ fig : whole ] ) shows the same spectra as in fig .
( [ fig : spectra2 ] ) in the broader photon energy range .
it is clear from fig .
( [ fig : whole ] ) that the reflected spectra in the `` thermal '' energy range ( e.g. , @xmath234 ev for the chosen x - ray flux ) , are not accurately computed ( as they never were in previous works on the x - ray reflection ) . in order to improve the treatment of the radiation field at these energies
, one would need to take into account stimulated emission and other many - body processes in the the cold layers of the illuminated gas , which we plan to do in our future work .
fortunately , spectra above @xmath235 kev should not be affected by this .
the critical information needed to understand the results presented herein is that for the relatively hard x - ray spectrum such as the one with @xmath194 , most of the iron atoms in the compton heated layer turn out to be nearly completely ionized , and thus both line emission and photo - absorption in this material are very weak .
we can define the following limiting cases to describe the behavior of the reprocessing features .
[ [ high - illumination - limit . ] ] high illumination limit .
+ + + + + + + + + + + + + + + + + + + + + + + + when the gravity parameter @xmath3 is small , i.e. , @xmath236 no easily visible reprocessing features result from the illuminated slabs .
the physics of this effect is rather transparent .
when the compton - heated layer is thomson thick , the x - rays incident on the disk atmosphere are reflected back before they can reach the cooler layers where the material is less ionized and where they could be absorbed and reprocessed into line or thermal radiation .
further , most of the photons will be reflected back by the thomson depth @xmath237 , which then means that the compton scattering changes the photon energy only slightly unless the photon is mildly relativistic ( @xmath238 kev ) .
therefore , it is to be expected that for photons of energy smaller than the energy corresponding to the peak in the neutral reflection component , i.e. , @xmath239 kev , the completely ionized material represents a perfect mirror ( only the photons angle changes after the scattering ) . above @xmath239 kev the photons are downscattered in a way very similar to that in the neutral reflection , so that the reflected spectrum above these energies is essentially unchanged ( see fig . [
fig : spectra2 ] ) .
[ [ low - illumination - limit . ] ] low illumination limit .
+ + + + + + + + + + + + + + + + + + + + + + + the low illumination case can also be called the `` strong gravity limit '' , when @xmath240 . as figure ( [ fig : temper1 ]
) shows , the thomson depth of the compton layer is much smaller than unity under these conditions , and hence its existence is of no noticeable consequence for the reflected spectra .
the latter are basically the same as the reflection spectra obtained in the classical studies of the neutral x - ray reflection as far as the medium to hard x - ray photons are concerned , ( e.g. , basko et al . 1974 ; lightman & white 1988 , magdziarz & zdziarski 1995 ) .
the iron line is rather narrow with centroid energy @xmath241 kev , and there is rather strong absorption edge above @xmath242 kev . [
[ moderate - illumination . ] ] moderate illumination .
+ + + + + + + + + + + + + + + + + + + + + + to a zeroth order approximation , the intermediate cases , when illumination is neither very strong or very weak , i.e. , @xmath243 , represent a combination of the low and the high illumination cases .
the thomson depth of the compton - heated layer is moderate ( @xmath244 ) , so the x - rays propagating in the downward direction have a fair chance to reach the cold layers .
the x - rays incident on the latter have spectrum similar to the incident one ( except for the high energy end of the spectrum ) , and therefore the spectrum that will be reflected off the cold layers ( i.e. , the one propagating in the upward direction at the interface between the compton and the cold layers ) will resemble the cold non - ionized reflection
. the compton scattering of the reflection hump " in the hot layers will not broaden it substantially .
however , the effects of compton scattering in the hot layers are quite pronounced for the energy range from about 10 @xmath245 to @xmath246 kev , because most photons in this energy band would be absorbed in the standard cold reprocessing ( see , e.g. , fig . 5 in magdziarz & zdziarski 1995 , or the lower solid curve in fig .
[ fig : whole ] ) .
the larger the thomson depth of the hot layer , the larger is the number of photons of this low energy ( @xmath247 kev ) band that are compton - reflected back within the _ hot upper layer _ , reducing the apparent strength of the reflection hump " .
one should note that the influence of the compton - heated layer on the x - ray reprocessed spectra was already discussed by basko , sunyaev & titarchuk ( 1974 ) , but this issue was not addressed at all in later works on x - ray reflection ( and , in fact , neither was the work of basko et al.(1974 ) ) .
it is worthwhile to point out that the iron line centroid energy remains at @xmath248 kev for all of our self - consistent calculations .
similarly , the position of iron edge remains the same .
these reprocessing features only get weaker , and thus less visible as the illumination increases ( @xmath3 decreases ) .
one can also note that there is a certain amount of broadening of the line by comptonization .
it is possible that this broadening can actually be detected if it is present in the data ( done 1999 , private communications ) . in order to accurately model line comptonization ,
it is desirable to perform higher resolution calculations using an exact klein - nishina scattering kernel , which we plan to do in the future .
if the illuminating flux is low , then the material is non - ionized in both self - consistent and constant density calculations , and its temperature is close to the effective one .
thus , the spectra in the low illumination limit discussed in [ sect : aares ] should be similar to those computed assuming constant gas density for @xmath249 or so ( e.g. , lightman & white 1988 ; @xmath7yckiet al . 1994 ; magdziarz & zdziarski 1995 ) .
likewise , if the illuminating flux is very high , the gas temperature is close to the compton one , and hence irrespectively of whether the gas density is fixed or found from hydrostatic equilibrium , the reflected spectra have very little iron line or edge .
this is the limit of the compton reflection , i.e. , the one without photo - absorption ( cf .
basko et al .
1974 , white et al . 1988 ) .
the reflected spectrum has virtually the same power - law index as the incident x - ray continuum , and only differs above @xmath250 few tens of kev .
figures ( [ fig : spectra1 ] ) and ( [ fig : spectra2 ] ) confirm these considerations . however , the differences in the reprocessing features between the self - consistent calculations with moderate values of the gravity parameter , i.e. , @xmath243 and the constant density calculations are truly remarkable ( see figs . [
fig : spectra1 ] and [ fig : spectra2 ] ) .
the former shows that the he - like iron line at @xmath251 and h - like line at @xmath252 kev never dominate over the `` cold '' iron line at @xmath253 kev , whereas the fixed density calculations predict that at @xmath254 the iron line at 6.7 kev will dominate .
in addition , under the constant density assumption , the equivalent width ( ew ) of the iron lines in the mildly ionized case can be larger by a factor @xmath255 compared with the line s ew in a neutral material ( see matt et al .
1993 ) . on the contrary ,
our calculations presented here show that solving for the pressure balance leads to the opposite behavior of the ew .
going from low illumination cases ( large @xmath3 ) to higher illumination ( lower @xmath3 ) , one finds that the strength of the iron lines rapidly decreases , and the absorption edge gets less noticeable ( cf .
[ fig : spectra1 ] ) . figure ( [ fig : angle1 ] ) shows the reflected spectra at three different viewing angles for the test with @xmath256 .
it is seen that the reprocessing features are strongest for the angle closest to the normal , and that they almost disappear at large angles .
the explanation for this effect is rather straight forward : the thomson depth of the compton heated layer , when viewed at an angle @xmath257 with respect to the normal is @xmath258 , where @xmath259 , and @xmath216 is the thomson depth of the compton layer as seen at angle @xmath260 .
further , since one only receives emission from material within one optical depth ( roughly speaking ) , at large inclination angles only the compton heated layer contributes to the reflected spectrum , and therefore no visible iron lines appear .
this effect should be important in modeling the iron lines from accretion disks that are seen almost edge - on . unless the optical depth of the completely ionized layer is very close to zero
, no iron line should be observed in such systems ( barring a more distant putative molecular torus ) .
figure ( [ fig : angle2 ] ) shows the temperature profile and the angle averaged spectra computed for @xmath261 , @xmath193 erg s@xmath39 @xmath40 , and @xmath219 for two different values of the x - ray incident angle @xmath262 .
the solid curve corresponds to @xmath263 ( i.e. , almost normal incidence ) , whereas the dotted curve corresponds to @xmath264 ( i.e. , almost parallel to the disk surface ) .
it is interesting to note that the temperature profiles are rather similar for moderate thomson depths .
we interpret this effect as due to the fact that the x - rays are isotropized by the time they penetrate to @xmath265 , so that the incident angular distribution does not matter . in the layer near the surface
, however , the difference is quite noticeable . in the case of the normal incidence ,
the gas temperature is only @xmath266 kev on the very top of the layer , while it is @xmath267 kev for the grazing angle of incidence .
the difference is easily explained by the fact that the compton temperature is approximately equal to @xmath268 where @xmath269 and @xmath270 are the _ angle integrated _ intensities of the x - ray and the reflected radiation ( e.g. , see begelman , mckee & shields 1983 ) . we should now recall that the x - ray flux @xmath19 is the flux projected on the direction normal to the disk .
therefore , @xmath271 , whereas @xmath272 .
note that @xmath273 , since the reflected radiation is clearly less isotropic than a truly isotropic field that would have @xmath274 and yet less beamed than a pencil beam escaping in the direction close to normal .
taking into account that @xmath275 , for x - rays that are normally incident on the disk , @xmath276 ; for grazing angles of incidence with @xmath277 , @xmath278 .
this is the reason why the gas temperature is hotter for larger incidence angles .
one can also note that this region with @xmath279 will not extend to very large @xmath66 , since at @xmath280 most of the incident x - rays will have been scattered at least once , which makes the x - radiation field more nearly isotropic .
this latter fact is clearly seen in figure ( [ fig : angle2]a ) .
turning to the difference in the reflected spectrum , figure ( [ fig : angle2]b ) demonstrates that the larger angles of incidence produce more hard x - ray continuum and the reprocessing features are weaker than in the case of nearly normal incidence .
this effect is due to a larger number of scatterings that the incident photons suffer before they reach the cool line - producing layers in the former versus the latter case ; i.e. when the x - rays come in at a large angle they are more likely to be reflected back before they can be photo - absorbed in the cold material . finally , we should mention that for this particular calculation , we turned off the x - ray pressure altogether , i.e. , assumed @xmath281 in equations ( [ heq1 ] ) and ( [ heq2 ] ) . as a result , the intermediate stable state ,
i.e. , the one with @xmath282 ev , occupies a considerably larger range of the thomson depth in fig .
( [ fig : angle2 ] ) than it did earlier for the tests shown in fig .
( [ fig : temper1 ] ) , when we self - consistently computed @xmath283 .
this therefore confirms that it is indeed the x - ray pressure that `` compresses '' the intermediate stable region ( see [ sect : struct ] ) compared to calculations of ko & kallman ( 1994 ) and r@xmath17aska(1999 ) .
figure ( [ fig : index ] ) shows how the temperature of the illuminated layer varies with the hardness of the incident x - rays while all the other parameters are held fixed . the fact that the maximum gas temperature increases with decreasing @xmath0 is easily understood because the compton temperature of a harder spectrum is higher .
in addition , the harder the spectrum , the larger is the optical depth of the compton heated layer , because it takes more scatterings to bring the average photon energy down such that the low temperature equilibrium states will become possible .
also note that for a steep spectrum , i.e. , for @xmath284 , the gas temperatures in the range @xmath285 ev become allowed ( because for such a spectrum the s - curve actually has a positive slope at these temperatures ) , whereas they are forbidden for harder spectra .
because of these differences in the temperature profile , it comes as no surprise that the reflected spectra are very different for hard and soft spectra . in figure ( [ fig : index2 ] ) we show the reprocessing spectra corresponding to the five tests shown in fig .
( [ fig : index ] ) . the hardest spectrum , i.e. , @xmath286 shows very little iron line emission .
it seems intuitively clear that when one adds this spectrum to the illuminating x - ray spectrum , the line and the reflection component will be barely discernible . for a soft x - ray spectrum , however , not only the 6.4
kev iron line is stronger , but there is a strong 6.7 kev line as well .
we now show several tests for a radiation - dominated accretion disk where the parameter @xmath3 was equal to its nominal value as defined by equation ( [ adef ] ) .
figure ( [ fig : rd ] ) shows the gas temperature profiles and the reprocessed spectra for two values of the incident x - ray flux ( @xmath287 erg @xmath40 s@xmath39 for the long - dashed curve , and @xmath288 erg @xmath40 s@xmath39 for the rest of the curves ) .
let us first concentrate on the long - dashed and the dotted curves .
these differ only by the absolute magnitude of the illuminating flux .
the self - consistent value of @xmath3 for these tests is @xmath289 and @xmath290 for the long - dashed and dotted curves , respectively . if we now compare these curves with the sequence of curves shown in the upper panel of fig .
( [ fig : temper1 ] ) , then we can observe that the radiation - dominated ( rd ) illuminated layers have much thicker compton - heated regions than fig .
( [ fig : temper1 ] ) would predict for similar values of @xmath3 .
the difference is due to the fact that the disk radiation force is strong even above @xmath179 for a rd disk : the radiation force due to the disk intrinsic emission cancels most of the gravitational force in the vicinity of the disk surface ( see equation [ heq2 ] and [ sect : pbc ] ) .
it is then convenient to introduce a `` modified gravity parameter '' @xmath291 defined as @xmath292 which then allows one to rewrite the equation for the hydrostatic balance for the illuminated layers above the rd disks as @xmath293\;. \label{heq3}\ ] ] in the simulations presented in figure ( [ fig : rd ] ) , the values of @xmath291 were @xmath294 and @xmath295 , for the dotted and long - dashed curves , respectively . in general , both terms ( @xmath291 and @xmath3 ) are important in establishing pressure balance , so that the classification of the reflected spectra by only one parameter @xmath3 is not as straight forward as it was for a gas - dominated disk ( [ sect : simple ] ) .
nevertheless , the behavior of the reflected spectra with @xmath3 is qualitatively the same . to investigate the importance of cooling due to the disk intrinsic flux
, we set up two additional calculations .
we kept @xmath291 and @xmath3 the same as for the dotted curve in figure ( [ fig : rd ] ) , but increased the intrinsic disk flux by factors 4 and 12 for the short - dashed and solid curves , respectively .
this test demonstrates that the usually unstable temperature region between @xmath94 ev and 1 kev can become thermally stable not only due to a soft x - ray incident spectrum ( see [ sect : indices ] ) , but also due to a large uv flux .
as one can deduce from equation ( [ tcom ] ) , the large intrinsic disk flux makes the maximum compton temperature equal to @xmath296 .
this means that the ionized layer may have a relatively low temperature .
when this temperature is below @xmath297 kev , the reflected spectrum will contain features characteristic of `` highly ionized '' reprocessing ( if the layer is somewhat optically thick ) , even though the ionizing spectrum is hard , in contrast to the results obtained for @xmath298 ( cf .
[ sect : indices ] ) .
so far we have been dealing with the problem of x - ray illumination in the context of a plane - parallel geometry where the x - rays come from infinity at a given incidence angle .
some of the possible geometries for the x - ray source and the accretion flow are not very far from this setup , while others may require solving a dynamical rather than a static problem .
for example , if the x - ray source is located at some height above the black hole ( `` lamp post model '' ) , as for example , in @xcite , then the reprocessed spectra can be obtained by our methods if one also takes into account the variation in the ionizing flux @xmath19 and the incident angle with radius .
one would then calculate the reflected x - ray spectra for a grid of radii and then integrate over the disk surface .
however , a difficulty that can not be resolved by our methods appears in the context of an accretion disk model where x - rays come from localized magnetic flares ( e.g. , galeev et al .
1979 ; field & rogers 1993 ; haardt et al .
1994 , nayakshin 1998c ; nayakshin & dove 1999 ) .
this difficulty is that the problem becomes not static because evaporation of the material into a _
wind _ may become important .
let us elaborate a little on this issue .
as we mentioned , the thermal gas pressure of the compton - heated layer can not drive a wind out to great distances because the compton temperature @xmath299 k is much below the virial temperature ( @xmath300 of the gas in the inner disk @xmath301 k ( see also begelman , mckee & shields 1983 ) .
similarly , the x - radiation for a sub - eddington accretion rates ( which we implicitly assumed here ) can not drive a global wind .
finally , line - driven winds ( e.g. , proga 1999 ; proga , stone & drew 1998 ) also can not remove the compton heated material since it is highly ionized and the line pressure force is basically zero .
what the x - radiation pressure and the thermal pressure can do is to move the compton - heated gas upwards or side - ways on distances of the order of few disk height scales @xmath163 ( one can check that moving the gas farther is counter - acted by the gravity and hence can not be done ) .
as far as we can see , this fact is of no consequence in the case of a single central x - ray source , whether it is located above the black hole or it is in the disk plane ( e.g. , fig .
2 in dove et al .
1997 ) , since the typical dimensions of the x - ray illuminated region are @xmath302 .
in contrast , moving the hot gas away from a magnetic flare , the source of the illuminating x - rays , on the distance of order few @xmath163 may be quite significant for the disks with flares , because most of the x - ray reprocessing will be done within this distance from the flare ( see fig . 1 in nayakshin & dove 1999 for an illustration ) .
once far enough from the flare , the gas will cool and settle down on the disk without much observational consequences . at the same time , the removal of the compton - heated gas away from the flares reduces the value of the thomson depth below the values calculated in this paper .
therefore , value of the gravity parameter @xmath3 alone does not define the reflected spectra in the case of a magnetic flare , and future calculations , including gas dynamics are needed .
fortunately , the reflected spectra depend most sensitively on the thomson depth , @xmath216 , of the compton layer between the source of x - rays and the cold material , irrespective of whether a particular value of @xmath216 is obtained in a static situation or in the local wind .
the wind velocity is comparable but somewhat larger than the thermal sound speed of the compton - heated gas ( see iii of mccray & hattchet 1975 in the context of a stellar x - ray induced wind ) , which is only @xmath303 of the speed of light . because the gas in the wind will probably be compton - heated , so that line opacities and emission are not important there , we do not expect any significant effects due the gas bulk motions on the radiative transfer .
accordingly , the x - ray reprocessing spectra should be very similar to the ones we obtained in this paper , covering the same low , moderate and high illumination limits , but the relation of these limits to the accretion disk and flare parameters ( e.g. , @xmath214 , @xmath19 , size of the flare , etc . )
is yet to be clarified .
we now summarize the important points about x - ray reprocessing and iron line emission emerging from the calculations presented in this paper .
we also discuss several observationally important issues where our methods should find useful applications .
the thermal ionization instability plays a crucial role in determining the density and temperature structure of the illuminated gas .
its effects can be taken into account by a proper treatment of the hydrostatic balance or dynamics of winds if the latter are important .
because of a ( nearly ) discontinuous behavior of the gas density and temperature caused by the thermal instability , one may not assume the illuminated gas to have either a constant or a gaussian density profile . for * hard x - ray spectra * with @xmath304 , and large x - ray flux ( i.e. , @xmath298 ) , the thermal instability `` forbids ''
the illuminated layers from attaining temperatures in the range from @xmath94 ev to @xmath305 typically @xmath118 few kev .
temperatures in the intermediate range @xmath306 ev are allowed , but due to the x - ray radiation pressure effects , the layers with these intermediate temperatures occupy a small thomson depth and thus contribute little to the reflected spectra . the major contributor to the reprocessing features ( lines , edges and
the characteristic reflection bump itself ) is the cool layer with temperature close to the effective temperature for the given x - ray and disk fluxes .
the overall structure of the illuminated layer is then approximately two - phase one ( neglecting the intermediate temperature layer ) : ( i ) the compton - heated `` skin '' on the top of the disk , and ( ii ) cold , dense , and thus typically weakly ionized layers below the hot skin .
the iron is completely ionized in the compton - heated material ( for hard spectra ) , and thus the only important radiative processes there are compton scattering and bremsstrahlung emission . as a result , the existence of the hot material always reduces the strength of the reprocessing features , because the incident x - rays may scatter back ( and out of the disk atmosphere ) before they reach the cold layers where atomic processes could imprint the characteristic reprocessing features . the thomson depth , @xmath216 , of the compton - heated illuminated layer
is controlled by the gravity parameter @xmath3 , whose values distinguish the following characteristic cases : @xmath307 low illumination , i.e. , high values of @xmath3 @xmath308 , which result in weakly ionized matter and lead to @xmath309 and hence to iron lines , edges and continuum reflection characteristic of neutral material .
@xmath307 moderate illumination , which corresponds to values of @xmath3 in the range @xmath310 , and produce moderately thick thomson - heated layers .
as far as the reflected continuum is concerned , the reprocessed spectrum represents a combination of that for the high and low illuminated cases discussed above .
this may not be true for the iron lines , though , because compton scattering of the line photons in the hot layer can be very effective in removing these from the line into the continuum ( done 1999 , private communications ) .
further work is needed in order to understand how this moderate illumination limit will appear when observed by a particular instrument .
we believe that because the fe atomic features are created in the cold nearly `` neutral '' material , and yet their normalization is lower than that of standard neutral reflection , some narrow band x - ray telescopes may confuse the mildly illuminated reflector covering a full @xmath61 solid angle with a _ non - ionized _ reflector covering a fraction of @xmath61 . @xmath307 high illumination ( @xmath311 ) .
x - radiation ionizes a substantial amount of material , so that the compton - heated layer is thomson - thick .
very little of either of the 6.4 , 6.7 or 6.9 kev iron lines are created in this layer because few x - ray make it through to the cold disk material .
thus , the reprocessing features , except for the roll - over at few tens kev due to the compton down - scattering are wiped out of the reflection spectrum .
the latter is a power - law with a similar index as the incident one for @xmath312 kev , and can be undetectable in the lower energy data .
@xmath307 the evolution of the reflection component and the iron lines from the weak illumination limit to the strong illumination limit is monotonic and in no point does the spectrum exhibit observational signatures of `` highly ionized '' matter ( as do the spectra of the constant density studies ) , because the compton layer is completely ionized and the line - creating material is very cold , effectively neutral .
this is to be contrasted with the predictions of constant density models , where the ew of the line and its centroid energy increases with ionization parameter @xmath30 ( e.g. , ross & fabian 1993 ; matt et al .
1993 ; @xmath7yckiet al .
1994 ; ross et al .
1999 ) .
ionized iron lines and strong absorption edges can nevertheless be produced under the following conditions : @xmath307 if @xmath224 , and the incident x - ray spectrum is steep , as in soft states of gbhcs , i.e. , @xmath304 , then the compton temperature even on the top of the reflecting layer can be lower than @xmath313 kev , which can then lead to the appearance of @xmath314 and @xmath315 kev iron lines .
@xmath307 if @xmath224 , and the incident x - ray spectrum is hard , @xmath316 , but the disk intrinsic flux exceeds the x - ray illuminating flux , i.e. , @xmath317 .
this again lowers the compton temperature of the hot layer and hence could yield ionized lines and strong absorption edges .
reflected spectra viewed at larger angles relative to the disk normal always contain less reprocessing features ( e.g. , lines and edges ) than those viewed face - on .
our findings are of interest for the modeling and the interpretation of observations from a number of astrophysical sources : recently , an x - ray `` baldwin '' effect was discovered by iwasawa & taniguchi ( 1993 ) in theginga data and with a much better statistics by nandra et al .
( 1997b ) in the _ asca _ data : the 6.4 kev fe line and the reflection component in the seyfert 1 x - ray spectra , which typically have @xmath318 kev luminosities @xmath319 a few @xmath320 ergs s@xmath39 can be well fit with reprocessing arising in low - ionization parameter material ( nandra et al .
higher luminosity sources exhibit a _ monotonic decrease _ of the fe line equivalent width ( ew ) ( nandra et al .
1997b ) . above @xmath321 ergs s@xmath39 , the ew of the line drops abruptly and this line becomes undetectable for @xmath322 ergs s@xmath39 , both in radio loud and radio quiet quasars .
these observations are consistent with the general trend of our calculations which indicate that upon increase of the x - ray flux , _ the x - ray reflecting material converts from `` cold '' to completely ionized _ , with a concurrent decrease in the strength of the fe 6.4 line , for the reasons discussed above . by contrast , models assuming a constant density for the illuminated gas predict that such an increase in the x - ray flux should give rise to the presence of intermediate states with even higher ew of he- and h - like iron lines and a deep broad iron edge ( see fig .
[ fig : spectra1 ] , and also fig . 1 in matt et al .
1993 ; fabian 1998 ) , which are not seen in the data .
our results show that , when computed self - consistently , i.e. , using hydrostatic pressure balance , the line ew and the depth of the edge monotonically decrease in accordance with the observations . in future work we plan to explore in detail the interpretation of this effect in the context of general , global agn models .
another application of our code is the modeling of the reprocessing features in transient bh sources , because these cover a wide range of accretion rates @xmath323 while they decline from the maximum luminosity . in studying the x - ray reprocessing features and their relativistic smearing
, one can potentially constrain such parameters of the accretion flow as the inner radius at which the disk terminates .
this kind of modeling was in fact performed by @xmath7ycki , done & smith ( 1997 ) , who studied reprocessing features in gs 2023 + 338 during its decline @xmath7yckiet al .
( 1998 ) who did the same for the nova muscae 1991 observations ; and by done & @xmath7ycki(1999 ) for cyg x-1 .
given the results of this paper , it seems rather likely that the x - ray reprocessing calculations that include hydrostatic balance can make a noticeable difference in terms of how close to the black hole the `` cold '' matter is in these sources , and what is the evolution of the flow geometry during the luminosity decline .
narrow line seyfert galaxies ( nlsg ) , that typically have steep x - ray spectra , may provide an additional opportunity to test accretion flows .
as we found , steep x - ray spectra may lead to a sufficiently low compton temperature , i.e. , @xmath324 kev , which then allows some of the highest ionization stages of iron ion to be abundant enough to imprint `` highly ionized '' signatures on the x - ray spectra .
in other words , due to steeper x - ray spectra , the hot compton layer may now become visible ( it is `` invisible '' for other objects with harder spectra because it does not produce any sharp line or absorption features ) , thus yielding interesting constraints on accretion flow geometry .
some of the observed nlsg indeed reveal unusual ionization states ( see , e.g. , comastri et al .
1998 and bautista & titarchuk 1999 ) .
in conclusion , we presented above a physically consistent way of computing the reprocessing features for any accretion disk geometry / model that does not produce x - ray driven winds . an important point to note
is that the reflected spectra depend mostly on the gravity parameter @xmath3 , which is determined by the illuminating flux , and the local gravity ( or the ratio @xmath215 ) .
the latter is independent of @xmath4-parameter the usual shakura - sunyaev viscosity parameter whose value is notoriously uncertain ( see , e.g. , nayakshin , rappaport & melia 1999 ) .
therefore , at least for a range of conditions ( to be specified in a future paper ) , the resulting reprocessed x - ray spectra are independent of @xmath4 and thus can be unambiguously calculated for a given accretion disk model .
therefore , we believe that the methods presented in this paper may be used as a meaningful and sensitive test of theoretical predictions versus observations of the iron lines , edges and the reprocessing features in general . with advent of the new x - ray telescopes ( chandra , astro - e and xmm ) , such tests may bring us one step closer to establishing which ( if any ) of the current accretion disk models is correct . sn acknowledges support by an nas / nrc associateship during the course of this work .
the authors thank manuel bautista , chris done , julian krolik , daniel proga , lev titarchuk and piotr @xmath7yckifor many useful discussions .
we can estimate the maximum thomson depth , @xmath325 , of the region where the heat conduction is important by the following considerations .
suppose that @xmath326 is the length scale over which the transition from the hot ( @xmath327 ) solution to the cold one ( @xmath328 ) occurs .
the maximum heat flux due to conduction is the saturated conductive flux , @xmath329 where @xmath330 is the isothermal sound speed at the hot ( compton ) solution , and @xmath28 is the gas pressure ( see cowie & mckee 1977 , and also max et al .
the x - ray heating per unit volume can be estimated as @xmath331 , where @xmath332 is the total cross section , and @xmath26 is the gas density .
therefore , heat conduction will dominates over distances no larger than @xmath333 , yielding the corresponding maximum thomson depth of @xmath334 where @xmath335 is the temperature of the hot zone adjacent to the temperature discontinuity in units of @xmath98 k ( note that @xmath326 is essentially the `` field length '' introduced by begelman & mckee ( 1990 ) when the conductive flux is given by the saturated flux ) .
here we made use of the fact that the transition from the hot to the cold solution happens when @xmath336 few@xmath39 .
therefore , the direct influence of heat conduction on the radiation transfer is negligible because the thomson depth of the region where conduction dominate is quite small ( similar conclusions were obtained by kmt in their appendix v ) .
we in fact can not afford to treat zones with such a small optical depth for numerical reasons , and thus taking thermal conduction into account is an expensive luxury for us . on the other hand
, heat conduction may be quite important in determining what solution the system picks in the unstable regime , but we believe that we have taken its effects into account as discussed in [ sect : multiplicity ] .
in the case of a thermally unstable medium , the cold and hot phases may actually be mixed , i.e. , the cold mater may be in the form of `` clouds '' embedded in the hot surrounding medium .
physics of cloud condensation and evaporation and radiative equilibria have been previously studied by , e.g. , zeldovich & pikelner 1969 ; cowie & mckee 1977 ; kmt ; balbus 1985 ; begelman & mckee and mckee & begelman ( 1990 ) .
although many aspects of the theory have been clarified in these papers , much uncertainty remains due to the unknown distribution of cloud sizes and separations .
r@xmath17aska & czerny ( 1996 ) , r@xmath17aska(1999 ) studied the role of the thermal instability in x - ray illuminated disks in agn .
they argued that instead of a sharp transition from the hot to the cold solution , there exists a complex two - phase medium with cold clouds embedded in a hotter gas .
however , the gas density of the cold phase is much larger ( by a factor of @xmath337 ) than that of the hot medium , so the cold clouds experience a much greater gravitational force than the hot medium , and thus they will sink , joining the cold ( one - phase ) medium below .
it seems quite likely that the gravity may separate the two phases , such that the hot phase will end up on the top of the cold phase just as a light oil does if poured into water , thus leading to a sharp boundary between the cold and the hot medium as suggested in our work here .
in addition , note that a cloud can also be broken up if its `` sinking velocity '' is even a small fraction of the sound speed , since then the face - on pressure is larger than the pressure on the sides of the cloud . in addition
, the incident x - ray flux may be variable , and thus can cause the height of the region where the two - phase medium may exist to be constantly changing , leading to cloud destruction before they can form . finally , even if some clouds do form and are efficiently maintained in a part of the illuminated layer , we believe that their existence is marginal for the radiation transfer and the resulting spectra .
indeed , in the case of the quasar emission line regions , the clouds are thought to form above the accretion disk , in the region filled with a hot compton - heated gas ( e.g. , krolik et al .
the latter , if optically thin , is basically invisible to the observer .
therefore , existence of any cool clouds embedded in this hot matter is of a significant observational interest , since the clouds are believed to be strong line emitters . in terms of a given spectral line , then , a cloud looks like a bright emitting object in the background of completely dark material .
the geometrical setup of the x - ray reprocessing problem is substantially different .
the cool clouds , if formed , exist in the region _ above _ the cold thermally stable gas .
moreover , the temperature of the clouds should be close to the temperature of the cold stable phase below .
this means that the emission spectrum of the clouds will be close to that of the cold material below .
thus , in terms of the observational signatures , except for increasing the amount of the cool material somewhat , the clouds will be completely washed out ( unresolvable ) on the background of the cold stable medium .
accordingly , we believe that existence of the two - phase medium in a region close to the temperature discontinuity is not only unclear , but is of academic value at least as far as the x - ray reprocessing is concerned .
it has been pointed out to us by julian krolik that while the viscous dissipation can be neglected in the energy equation of the illuminated layer , one may still have to consider the effects of the viscous heating for hydrostatic balance ( equation [ heq1 ] ) .
this is so because the gravity is canceled mainly by the radiation pressure even in the illuminated layer of a radiation - dominated disk ( see [ sect : rd ] ) .
thus , there could be a situation in which the increase with height in the first term on the right in equations ( [ heq1 ] , [ heq2 ] ) is entirely balanced by a corresponding increase in the outward radiation pressure force ( the third term ) , which is exactly the case in a rd ss73 disk .
one can show that for this to be true , the gas density of the illuminated layer must stay close to its midplane value ( because viscous dissipation is usually modeled as being proportional to the gas density
see ss73 ) , a rather unlikely condition for the illuminated matter in the setup of our paper .
it is furthermore impossible if one resolves the vertical disk structure more accurately .
simulations ( agol and krolik 1999 , private communications ) show that convection changes the disk profile in a way such that the gas density does not remain constant inside the rd disks ; the gas density is greatest in the midplane and it decreases with height . | we study the illumination of accretion disks in the vicinity of compact objects by an overlying x - ray source .
our approach differs from previous works of the subject in that we relax the simplifying assumption of constant gas density used in these studies ; instead we determine the density from hydrostatic balance which is solved simultaneously with the ionization balance and the radiative transfer in a plane - parallel geometry .
we calculate the temperature profile of the illuminated layer and the reprocessed x - ray spectra for a range of physical conditions , values of photon index @xmath0 for the illuminating radiation , and the incident and viewing angles . in accordance with some earlier studies , we find that the self - consistent density determination makes evident the presence of a thermal ionization instability well known in the context of quasar emission line studies .
the main effect of this instability is to prevent the illuminated gas from attaining temperatures at which the gas is unstable to thermal perturbations .
thus , in sharp contrast to the constant density calculations that predict a continuous and rather smooth variation of the gas temperature in the illuminated material , we find that the temperature profile consists of several well defined thermally stable layers . transitions between these stable layers are very sharp and can be treated as discontinuities as far as the reprocessed spectra are concerned . in particular , the uppermost layers of the x - ray illuminated gas are found to be almost completely ionized and at the local compton temperature ( @xmath1 k ) ; at larger depths , the gas temperature drops abruptly to form a thin layer with @xmath2 k , while at yet larger depths it decreases sharply to the disk effective temperature . for a given x - ray spectral index , this discontinuous temperature structure
is governed by just one parameter , @xmath3 , which characterizes the strength of the gravitational force relative to the incident x - ray flux .
we find that most of the fe k@xmath4 line emission and absorption edge are produced in the coolest , deepest layers , while the fe atoms in the hottest , uppermost layers are generally almost fully ionized , hence making a negligible contribution to reprocessing features in @xmath5 kev energy range .
we also find that the thomson depth of the top hot layers is pivotal in determining the fraction of the x - ray flux which penetrates to the deeper cooler layers , thereby affecting directly the strength of the fe line , edge and reflection features . due to the interplay of these effects , for @xmath6 , the equivalent width ( ew ) of the fe features decreases monotonically with the magnitude of the illuminating flux ,
while the line centroid energy remains at 6.4 kev .
we provide a summary of the dependence of the reprocessing features in the x - ray reflected spectra on the gravity parameter @xmath3 , the spectral index @xmath0 and other parameters of the problem .
we emphasis that the results of our self - consistent calculations are both quantitatively and qualitatively different from those obtained using the constant density assumption .
therefore , we propose that future x - ray reflection calculations should always utilize hydrostatic balance in order to provide a reliable interpretation of x - ray spectra of agn and gbhcs . |
a long standing open problem in the mathematical analysis of the stochastic ising model @xcite can be described as follows .
consider the standard @xmath12 spin ising model at inverse temperature @xmath13 and zero external magnetic field in a cubic box @xmath14 , @xmath15 , of side @xmath2 , with homogeneous , e.g. `` plus '' , boundary conditions outside @xmath16 ( see section [ sec : model ] for the precise definition ) .
denote by @xmath17 the corresponding gibbs measure and assume that @xmath13 is larger ( or much larger ) than the critical value @xmath18 where the `` plus '' and the `` minus '' phases start to coexist . on the spin configuration space @xmath19
consider also the continuous time markov chain , reversible w.r.t .
the gibbs measure @xmath17 , in which each spin @xmath20 , @xmath21 , with rate one chooses a new value from @xmath12 with probabilities given by the conditional gibbs measure given the current values of the spins outside the site @xmath22 .
such a chain is known in the literature as _ gibbs sampler _ or _ glauber chain _ and , because of reversibility , its unique stationary distribution coincides with @xmath17 .
let @xmath23 denote the distribution of the chain at time @xmath24 when the initial configuration at time @xmath25 is @xmath26 and let @xmath27 be the minimum time such that the variation distance between @xmath23 and @xmath17 is smaller than , e.g. , @xmath28 for any @xmath29 . because @xmath16 is a finite set and the chain is ergodic , the perron - frobenius theorem guarantees that @xmath27 is finite and it is not too difficult to prove that @xmath30 for some constant @xmath31 ( see e.g. @xcite ) .
despite the fact that the above upper bound is known to be the correct order of growth of @xmath27 when the boundary conditions are absent , i.e. free ( see e.g. @xcite ) , it is known that the presence of the homogeneous `` plus '' boundary conditions drastically modifies the whole process of relaxation to equilibrium . in this case the conjectured correct growth of @xmath27 as a function of @xmath2 is given by the lifshitz s law @xmath32 ( see @xcite and @xcite ) .
unfortunately , _ any _ polynomial upper bound on @xmath33 has escaped so far a rigorous analysis and the best known result , following a recent breakthrough @xcite , is confined to the two dimensional case @xmath9 and it is of the form @xmath34 for any @xmath35 ( see @xcite ) .
the conjectured @xmath10 growth of @xmath27 is related to the shrinking of a `` spherical '' bubble of the `` minus '' phase under the influence of the `` plus '' boundary conditions . on a macroscopic scale one expects @xcite that the dynamics of the bubble follows a _ motion by mean curvature _ which , if true , implies immediately that in a time @xmath36 the bubble disappears and equilibrium is achieved .
some of the above questions and in particular the validity of the lifshitz s law can be formulated ( and its mathematical justification remains highly non - trivial ) even at zero temperature ( @xmath37 ) , a situation that was considered in great detail in @xcite ( see also @xcite , @xcite ) .
similarly one can consider , as we do in the sequel , an almost zero temperature , i.e. @xmath38 and increasing so fast with @xmath2 that thermal fluctuations become irrelevant on the relaxation process . in the extreme case @xmath37
, the stationary distribution is concentrated on the `` plus '' configuration ( i.e. all spins equal to @xmath39 ) and the glauber chain evolves towards it without ever increasing the spin energy .
the mixing time @xmath27 becomes closely related to the hitting time of the `` plus '' configuration starting from all minuses and it is not too hard to prove , by induction on the dimension @xmath40 ( starting from the case @xmath9 , which requires non - trivial work @xcite ) , an upper bound of the form @xmath41 .
however the inductive argument , which in @xmath1 simply boils down to comparing the true evolution of the original cubic bubble of minuses with an auxiliary chain in which the @xmath2 two - dimensional square layers of the bubble disappear one after the other starting e.g. from the top one , completely neglects the interesting cooperative effect in which the whole bubble , starting from the corners , is eroded by a `` mean curvature effect '' which enhances considerably its shrinking . in other words , the lifshitz s law in @xmath1 can not be proved or even approached closely without considering in detail how a two - dimensional curved interface separating the pluses from the minuses evolves in time .
the first main contribution of the present work is a strategy to attack and solve the above problem at zero temperature for @xmath1 ( with logarithmic corrections ) and for @xmath9 ( with different constants in the upper and lower bounds ) .
we refer to section [ sec : results ] for the precise statements . in the challenging three dimensional case
our method involves the use and adaptation to our specific situation of two different and beautiful sets of results concerning : a. the gaussian free field - like equilibrium fluctuations of _ random monotone surfaces _ and their connection with _ random dimer coverings _ of the two - dimensional hexagonal lattice ( see @xcite ) ; b. the mixing time of a glauber chain for monotone surfaces @xcite . the key point of our approach is to prove that the evolution of the glauber chain is dominated by that of another effective chain which follows a motion by mean curvature , only slowed down by a logarithmic factor ( in the radius of the bubble ) . in turn
the evolution of the bounding chain is constructed by `` peeling off '' at time @xmath24 a layer of logarithmic width from the bubble of radius @xmath42 .
the peeling process is realized by letting relax to equilibrium a mesoscopic spherical cap of radius @xmath43 and height @xmath44 ( where @xmath45 stands for a suitable polynomial of @xmath46 ) centered at each point of the surface of the bubble .
results ( i ) above are essential in order to prove the peeling effect while the results ( ii ) prove that the overall effect occurs on the correct time scale @xmath47 .
we strongly believe that our approach can be helpful in solving other related problems . the second contribution , this time for the model in dimension @xmath9 , are upper and lower bounds , _ linear _ in @xmath48 and _ uniform _ in @xmath49 , @xmath31 large enough , on the _ spectral gap _
( see ) of the glauber chain . in @xcite for @xmath9
the upper bound was proved ( apart from logarithmic corrections and with constants depending on @xmath13 ) for any @xmath35 and it was conjectured to be the correct behavior of the spectral gap ( to be compared with the @xmath10 scaling of the mixing time ) . in our case the proof of the lower bound ( the most interesting one ) has an analytic flavor .
we first unitarily transform the original markov generator of the glauber chain acting on @xmath50 into a new matrix acting on @xmath51 with the flat ( i.e. counting ) measure , whose off - diagonal elements corresponding to spin transitions which _ do not conserve the energy _ vanish very fast as @xmath52 .
such a property , for @xmath53 , allows us to write the new matrix into a block - form plus a remainder whose norm is very small with @xmath2 . since each block describes a suitably killed glauber chain ( killing occurs as soon as the spin energy decreases ) the desired bound follows by a probabilistic analysis of the killing time for each block .
given @xmath54 , we write @xmath55 if @xmath22 is a nearest neighbor of @xmath56 . to each point
@xmath57 is assigned a spin @xmath20 which takes values in @xmath58 .
given @xmath14 , we let @xmath59 and @xmath60 be the set of all possible configurations in @xmath16 . the gibbs measure in a finite domain @xmath16 for a boundary condition @xmath61 at inverse temperature @xmath62 is given by @xmath63 with @xmath64 and @xmath65 when no danger of confusion arises , we will write just @xmath66 for @xmath67 . we will consider also the case @xmath37 , in which case @xmath67 is taken to be the uniform measure over all configurations @xmath68 which minimize @xmath69 .
the dynamics we consider is the glauber dynamics @xmath70 , where @xmath71 is the initial condition . to each @xmath72
is associated an independent poisson clock of rate @xmath73 .
when the clock labeled @xmath22 rings , say at time @xmath74 , one replaces @xmath20 with a value sampled from the probability distribution @xmath75 , where @xmath76 we denote the law of @xmath77 for a given time @xmath24 as @xmath78 .
it is well known that @xmath70 is a markov process , reversible with respect to the equilibrium measure @xmath66 .
we are interested in the order of magnitude of the `` mixing time '' , defined for some @xmath79 as @xmath80 where @xmath81 denotes the total variation distance between two probability measures . when @xmath82 , we will just write @xmath83 for @xmath84 .
another quantity we will focus on is the spectral gap , @xmath85 where @xmath86 denotes the variance w.r.t .
@xmath67 , @xmath87 is the generator of the dynamics , @xmath88 ( a more explicit expression for the generator is given in section [ sec : lbgap3d ] ) and the infimum is taken over non - constant functions @xmath89 .
the inverse of the gap ( respectively the mixing time ) measures the speed of convergence to equilibrium in @xmath90 ( resp . in total variation norm ) .
also , it is well known that @xmath91 which actually holds for any reversible markov process . for notational clarity , in this subsection
we indicate explicitly the dependence on the boundary condition @xmath92 in @xmath93 , the law of the dynamics @xmath94 at time @xmath24 , @xmath95 being the initial condition .
the glauber dynamics enjoys the following well known monotonicity property .
one can introduce a partial order in @xmath19 by saying that @xmath96 if @xmath97 for every @xmath72 .
then , one has @xmath98 where @xmath99 denotes stochastic domination ( one writes @xmath100 if @xmath101 for every increasing function @xmath102 , i.e. @xmath103 whenever @xmath104
; an event will be called increasing if its characteristic function is increasing ) . in particular , letting @xmath105 one obtains @xmath106 also , it is possible to realize on the same probability space the trajectories of the markov chain corresponding to distinct initial conditions @xmath95 and/or distinct boundary conditions @xmath92 , in such a way that , with probability one , @xmath107 this will be referred to as the `` monotone global coupling '' : its law will be denoted by @xmath108 and the corresponding expectation by @xmath109 . throughout the paper we will apply several times the above monotonicity properties : for brevity , we will simply say `` by monotonicity ... '' instead of referring explicitly to - . in this section , it is understood that @xmath16 is the cubic domain @xmath111 of side @xmath112 and that @xmath113 for every @xmath114 , so that we omit @xmath16 and @xmath92 from the notations .
in this work , we consider the situation where @xmath13 grows with @xmath2 , and in particular @xmath115\end{aligned}\ ] ] for a sufficiently large @xmath116 . note that , for @xmath37 , the equilibrium measure concentrates on the all `` @xmath0 '' configuration : @xmath117 .
given @xmath7 , one can choose @xmath116 large enough so that @xmath118 for every @xmath13 in the range ( this follows from easy peierls estimates ) .
moreover , for every initial condition @xmath119 , one can find a coupling between the dynamics at @xmath37 and @xmath120 such that they coincide until time @xmath121 with probability at least @xmath122 ( just compare the transition rates ) .
in particular , this implies that @xmath123 we will see later ( cf .
remark [ rem : dinuguali ] ) that , as an immediate consequence of theorem [ th : d3 ] , one actually has the estimate @xmath124 we define the random time @xmath4 as the first time when all spins are `` @xmath0 '' , starting from the all `` @xmath3 '' configuration : @xmath125 observe that , under the global coupling , if @xmath37 , then @xmath126 for every @xmath127 and for every initial condition @xmath95 .
* notational conventions * * for notational clarity , quantities referring to the @xmath37 dynamics will have a superscript @xmath128 ( we will write for instance @xmath129 and @xmath130 ) , while ( for lightness of notation ) we will not put a superscript @xmath13 when @xmath131 ; * our main focus will be on the case of the all `` @xmath0 '' boundary condition ( @xmath132 for every @xmath133 ) and we will just write @xmath134 in this case ; * we let @xmath135 denote the law of the process @xmath136 at @xmath37 , started from the `` @xmath3 '' configuration , with boundary condition @xmath134 ; * given @xmath68 and @xmath72 , we will denote by @xmath137 the configuration obtained by flipping the spin at @xmath22 to @xmath138 .
based on the observations and , one can use the time @xmath4 to estimate the mixing time for @xmath13 large : [ th : t+tmix ] for every @xmath79 one has @xmath139 where we recall that @xmath135 is the law of the process @xmath136 for @xmath37 , with `` @xmath0 '' boundary conditions .
moreover , fix @xmath79 and @xmath140 and assume that , for some @xmath141 , @xmath142 for @xmath2 sufficiently large . if @xmath143 with @xmath116 sufficiently large , then for every @xmath2 large one has @xmath144 and @xmath145 _ proof of lemma [ th : t+tmix ] _ . recall that @xmath146 so that @xmath147 ( where the inequality follows from monotonicity ) and is immediate . to prove - , it is sufficient to choose @xmath116 sufficiently large so that and hold for @xmath148 and to take @xmath2 sufficiently large so that @xmath149 .
consider the 3d ising model in the cubic box @xmath151 with boundary condition @xmath152 and @xmath37 . the main result of this work is : [ th : d3 ] there exists a positive constant @xmath153 such that for @xmath154 one has @xmath155\ge 1-\frac cl .
\end{aligned}\ ] ] [ rem : dinuguali ] 1 . as we mentioned in the introduction ,
the previously known upper bound for @xmath4 was of order @xmath156 ( * ? ? ? * th .
1.3 ) , while matches the heuristically expected behavior , except for the logarithmic factor . on the other hand
, we are not aware of previously known lower bounds , except for the trivial estimate @xmath157 ; 2 . via lemma
[ th : t+tmix ] , once we prove we also have that @xmath158 for @xmath159 and @xmath116 large ; 3 . to prove , observe first of all that @xmath160 the first two terms are estimated through , taking say @xmath161 and using the upper bound for the mixing time .
the proof is concluded thanks to - ( if @xmath116 is chosen such that these bounds hold for @xmath162 ) .
for the 2d ising model in the cubic box @xmath163 with boundary condition @xmath152 and @xmath37 we have [ th : d2 ] there exist positive constants @xmath164 such that @xmath165 for every @xmath166 .
[ rem : d2tmix ] thanks to lemma [ th : t+tmix ] , this implies that @xmath167 for @xmath159 .
we mention that the bound @xmath168 was proven in ( * ? ? ?
1.3 ) ; as for the lower bound for @xmath4 , the authors of @xcite proved the following weaker result : for every @xmath140 , @xmath169 tends to zero as @xmath170 . for a simplified version of the dynamics
, the authors of @xcite proved the sharp behavior @xmath171 . in two dimensions , always with `` @xmath0 '' boundary conditions , we also have sharp bounds on the spectral gap : [ th : d2gap ] there exist positive constants @xmath172 such that for every @xmath173 , one has @xmath174
the reason why we require @xmath131 is that for @xmath37 the equilibrium measure is concentrated on a single configuration ( the all `` @xmath0 '' configuration ) , and the spectral gap is not well defined ( all functions have zero variance with respect to @xmath66 ) .
[ rem : fss ] on the basis of the analysis of a one - dimensional toy model of evolution of a bubble of `` @xmath3 '' phase inside the `` @xmath0 '' phase , it was conjectured in ( * ? ? ?
7 ) that in dimension @xmath9 the spectral gap behaves like @xmath175 for every @xmath35 .
our result is the first confirmation of this conjecture , in the @xmath176 limit .
let us remark also that the authors of @xcite showed , via the construction of a suitable test function , that @xmath177 for every @xmath35 .
however , when @xmath13 grows with @xmath2 as in , the constants @xmath178 possibly diverge .
therefore , even the upper bound in is a qualitative improvement over known results . the proof that @xmath179 for @xmath159 extends easily to the three - dimensional model ( cf . section [ sec : lbgap3d ] )
but , as we discuss in a moment , there is no reason to believe that this is the correct behavior of the spectral gap in @xmath1 .
two interesting questions which are left open by the above results are the following : 1 . what is the order of magnitude of the spectral gap in dimension @xmath1 ? one expects it to be much larger than in dimension @xmath9 , possibly of order @xmath180 for some @xmath7 , or even of order @xmath73 ( * ? ? ?
7 ) , @xcite ; 2 . how to remove the logarithmic corrections in the upper and lower bounds for @xmath83 in dimension @xmath1 ? a suggestive indication that indeed @xmath181 comes from the following argument .
if one defines the `` entropy constant '' ( or modified log - sobolev constant ) as @xmath182 where the supremum is taken over positive functions @xmath102 which verify @xmath183 , it is possible to exhibit a test function which gives @xmath184 for some constant @xmath31 which is positive , uniformly for @xmath185 with @xmath116 large . on the other hand
, it has been conjectured @xcite ( and verified in various explicitly solvable examples ) that for every reversible markov process the mixing time is lower bounded by @xmath186 times some universal constant .
the organization of the paper is as follows .
theorem [ th : d3 ] is proven in section [ sec : ubmt ] ( upper bound on @xmath4 ) and in section [ sec : lbmt ] ( lower bound ) .
the upper bound requires precise estimates on the fluctuations of dimer coverings of the hexagonal ( or honeycomb ) lattice , see section [ app : kenyon ] , and on the mixing time of a dynamics on plane partitions or monotone sets , proven in section [ app : wilson ] .
as for the 2d model , theorem [ th : d2 ] is proven in section [ sec : tm2d ] and theorem [ th : d2gap ] in section [ sec : gap2d ] . * more notational conventions * * in the proof of the results , @xmath187 etc .
denote positive and finite constants ( independent of @xmath2 and @xmath13 ) which are not necessarily the same at each occurrence . *
if @xmath188 , it will be understood that @xmath189 are its cartesian coordinates .
* @xmath190 will denote the euclidean distance in @xmath191 .
* @xmath192 will denote the geometric boundary of a set @xmath193 , except when @xmath194 , in which case @xmath195 .
* @xmath196 denotes the set of non - zero integers @xmath197 .
[ def : monset ] we say that @xmath198 is a _ positive monotone set _
if @xmath199 implies @xmath200 whenever @xmath201 and @xmath202 .
the collection of all positive monotone sets will be denoted by @xmath203 , which is partially ordered with respect to inclusion .
when @xmath204 is a finite set , it is usually called a _ plane partition _ , the two - dimensional generalization of an ordinary partition ( or _ young diagram _ ) . given @xmath205 in @xmath203 such that @xmath206
, we define a dynamics @xmath207 on @xmath203 such that @xmath208 for all times , with initial condition @xmath209 .
let @xmath210 and associate to each @xmath211 an independent poisson clock of rate one .
when the clock labeled @xmath212 rings at some time @xmath74 : * if both sets @xmath213 and @xmath214 belong to @xmath203 , then we replace @xmath215 with @xmath216 , with @xmath217 chosen between `` @xmath0 '' and `` @xmath3 '' with equal probabilities @xmath218 ; * otherwise , we keep @xmath215 unchanged .
the link with the @xmath37 dynamics of the ising model is straightforward : consider the ising dynamics in the domain @xmath219 , with initial condition @xmath220 and with boundary conditions @xmath221 and let @xmath222 .
then , @xmath223 has the same law as @xmath224 . to avoid any risk of confusion we choose different notations for the ising dynamics and the monotone set dynamics : we call @xmath225 the law of @xmath224 and @xmath226 its ( reversible ) invariant measure .
of course , @xmath227 is just the uniform measure over the positive monotone sets @xmath204 such that @xmath228 .
the monotonicity properties of the glauber dynamics discussed in section [ sec : monotone ] immediately give : [ th : mondyn ] if @xmath229 and @xmath230 , then @xmath231 .
the next key result quantifies the mixing time of the monotone set dynamics as a function of the shape of @xmath16 : [ th : wilson ] let @xmath232 and @xmath233 if @xmath234 , the mixing time of the monotone set dynamics is @xmath235 we believe that the correct behavior is @xmath236 .
the spurious factor @xmath237 is potentially dangerous , so we will take care to apply theorem [ th : wilson ] only in situations where @xmath238 is small , say of order of a power of @xmath239 : this is a crucial step for the proof of the upper bound , which differs from the expected behavior @xmath36 only by logarithmic corrections .
we give the proof of theorem [ th : wilson ] in section [ app : wilson ] .
let @xmath240 where @xmath241 is the ball of radius @xmath242 centered at the origin . by monotonicity
, the claim follows if we prove the following result for the dynamics in @xmath243 with `` @xmath0 '' boundary conditions , started from `` @xmath3 '' : [ th : sfere ] consider the dynamics in @xmath243 with `` @xmath0 '' boundary conditions on @xmath244 .
there exists @xmath7 such that for every @xmath245 and @xmath246 one has @xmath247 the reason why this result implies is that the set @xmath248 contains a cube , call it @xmath249 , which is a translate of @xmath250 and , by monotonicity , the marginal in @xmath249 of the evolution in @xmath243 is stochastically dominated by the evolution in @xmath249 started from `` @xmath3 '' , with `` @xmath0 '' boundary conditions on @xmath251 . in order to get proposition [ th : sfere ] , we prove [ th : sfere2 ] consider the dynamics in @xmath252 with `` @xmath0 '' boundary conditions and let @xmath253 .
there exists @xmath7 such that for every @xmath245 @xmath254\right)\ge 1-c / l^{4}. \end{aligned}\ ] ] proposition [ th : sfere ] then easily follows : _ proof of proposition [ th : sfere ] ( assuming proposition [ th : sfere2 ] ) _ by the union bound , one then deduces that @xmath255\mbox{\ ; such that \ ; } \sigma^-_x(t)=-\right ) \le c / l^{2}.\end{aligned}\ ] ] for the dynamics in @xmath243 we prove , by induction on @xmath256 , that @xmath257\mbox{\ ; such that\ ; } \sigma^-_x(t)=- \right)\le \frac { c\,i}{l^2},\end{aligned}\ ] ] which for @xmath258 implies . for @xmath259 ,
this follows from eq . .
if the claim holds for some @xmath260 , then @xmath261 where for any event @xmath262 we let @xmath263 denote its complement .
next , by monotonicity , @xmath264 \;:\ ; \hat \sigma_x(t)=-\right)\end{aligned}\ ] ] where @xmath265 is the evolution in @xmath266 with `` @xmath0 '' boundary conditions , which starts from `` @xmath3 '' at time @xmath267 .
thanks to monotonicity we can restart from `` @xmath3 '' the evolution at time @xmath267 and we used the hypothesis @xmath268 to freeze all spins outside @xmath269 to the value `` @xmath0 '' in the time interval @xmath270 $ ] . via a trivial time translation , the upper bound ( applied with @xmath2 replaced by @xmath271 ) shows that the right - hand side of is smaller than @xmath272 ( @xmath31 being the same constant which appears in ) which , together with , completes the inductive proof .
@xmath273 _ proof of proposition [ th : sfere2 ] .
_ clearly , it is sufficient to prove the claim for @xmath2 large enough . in the following
, @xmath274 will denote a small positive universal constant ( conditions on its smallness will be specified later ) .
given a point @xmath275 , let @xmath276 be its spherical coordinates , where @xmath277 $ ] is the polar angle and @xmath278 is the azimuthal angle , with the convention that the half - plane @xmath279 corresponds to @xmath280 and that @xmath281 identifies the positive @xmath282 axis .
note that the portion of @xmath283 contained in the octant @xmath284 is a monotone surface , where : [ def : monotone ] a smooth surface @xmath285 is said to be monotone if for every @xmath286 all three components of the normal vector @xmath287 at @xmath22 are non - zero and have the same sign .
it is convenient to introduce a few geometric definitions : [ def : geom ] for a given @xmath288 , let a. @xmath289 be the infinite half - line which starts from the origin of @xmath290 and goes through @xmath22 b. @xmath291 be the intersection of @xmath289 and @xmath292 c. @xmath293 be the plane perpendicular to @xmath289 which meets @xmath289 at distance @xmath294 from the origin ( the exponent @xmath295 is somewhat arbitrary , and it could be replaced by @xmath296 for any @xmath140 ) d. @xmath297 be the half - space not containing the origin and delimited by @xmath293 e. @xmath298 be the spherical cap @xmath299 .
also , decompose the boundary of @xmath298 as @xmath300 , where @xmath301 is the `` curved portion of the boundary '' and @xmath302 is a disk . thanks to the discrete symmetries of the sphere @xmath303 , it is clearly enough to prove for points @xmath253 such that @xmath304 $ ] and @xmath305 $ ] .
actually , we claim that it is enough to restrict to @xmath306 $ ] and @xmath305 $ ] , if @xmath140 is small enough . indeed ,
if we interchange the role of the first and third coordinates of @xmath290 , the points with @xmath305 $ ] and @xmath307 close to @xmath308 are mapped into points with @xmath309 and @xmath310 small .
it is convenient to distinguish three cases ( see figure [ fig : sfera ] ) : of the surface of the sphere of radius @xmath2 . by symmetry
, we need to prove only in the case where @xmath311 and @xmath291 ( cf
. definition [ def : geom](b ) ) does not fall in the strip of width @xmath312 , adjacent to the @xmath313 plane .
case * a * corresponds to @xmath291 in @xmath314 ( the subset of the surface of the sphere delimited by the thick line ) .
case * b * corresponds to @xmath291 at geodesic distance at most @xmath315 from the north pole .
finally , case * c * corresponds to @xmath291 in the strip of width @xmath316 to the left of @xmath314 . for reasons of graphical clarity
, proportions are not respected in the drawing .
, scaledwidth=60.0% ] 1 .
* case a : * @xmath253 is such that @xmath317 where @xmath318\times[0,\pi/4]\right)\cap\left\{(\theta,\phi):\theta\ge \frac{\log l}{\sqrt{l}},\;\phi\sin(\theta)\ge \frac{(1+\delta)\sqrt2}{\sqrt l}(\log l)^{3/4}\right\ } . \end{aligned}\ ] ] remark that the lower bound on @xmath319 just means that , moving on @xmath283 along a line of constant @xmath307 , the distance between @xmath291 ( cf .
definition [ def : geom](b ) ) and the set of points in @xmath283 where @xmath280 is at least @xmath320 .
+ under the condition @xmath321 , one has : [ th : geomovvio ] 1 .
the disk @xmath322 has radius @xmath323 2 .
@xmath324 is contained in the monotone octant @xmath325 . + as a consequence , both @xmath324 and @xmath322 are monotone surfaces , cf .
definition [ def : monotone ] .
+ _ proof of lemma [ th : geomovvio ] . _
statement ( a ) and monotonicity of @xmath322 require just elementary geometric considerations . as for statement
( b ) , it is easy to see that the condition @xmath326 implies that the geodesic distance of @xmath291 along @xmath327 from the boundary of @xmath328 is larger than @xmath329 . the fact that @xmath324 is contained ( for @xmath2 large enough ) in @xmath330 then just follows ( thanks to the fact that @xmath274 is strictly positive ) from statement ( a ) .
+ @xmath331 + we can now continue the proof of proposition [ th : sfere2 ] under the condition @xmath321 .
let @xmath332 be the collection of lattice sites contained in @xmath298 ( cf .
definition [ def : geom](e ) ) .
it is clear that the site @xmath253 under consideration belongs to @xmath333 and that , thanks to lemma [ th : geomovvio](a ) , @xmath334 by monotonicity , it is enough to prove for a modified dynamics where only the spins @xmath335 evolve , starting from the `` @xmath3 '' configuration , with boundary conditions given by @xmath336 for @xmath337 and @xmath338 for @xmath339 . for lightness of notation , we still call such dynamics @xmath340 .
let @xmath341 and observe that @xmath342 are positive monotone sets in the sense of definition [ def : monset ] , as guaranteed by the monotonicity of the surfaces @xmath343 ( cf .
lemma [ th : geomovvio ] ) . also , call @xmath344 : thanks to the discussion in section [ sec : wilson ] , we have that @xmath223 is a positive monotone set at all times @xmath345 , and trivially @xmath346 . to be coherent with the notations of section [ sec : wilson ] , we have in the present case @xmath347 , @xmath348 ; also , the law of @xmath223 is just @xmath225 , with invariant measure @xmath349 . finally note that , from the definition of the spherical cap @xmath298 , one has @xmath350 ( recall that we are working under the assumption that @xmath351 ) .
putting together theorem [ th : wilson ] with the estimates , , we get that the mixing time of the dynamics in @xmath333 is @xmath352 . +
we have the following key equilibrium estimate : + [ th : kenyon ] recall that @xmath353 is the uniform distribution over all positive monotone sets @xmath204 such that @xmath228 .
there exists @xmath7 such that for every @xmath154 and whenever @xmath354 , @xmath355\le \frac1c\exp(-c(\log l)^{3/2 } ) .
\end{aligned}\ ] ] + in words , this result is saying that spins which are at distance of order @xmath356 away from the bottom of the spherical cap @xmath333 , where the `` @xmath3 '' boundary conditions act , are `` @xmath0 '' with overwhelming probability .
it is crucial here that the estimate ( i.e. the value of @xmath31 ) does not depend on the spherical coordinates @xmath357 of the point @xmath22 under consideration , i.e. on the slope of the plane @xmath358 of definition [ def : geom](c ) .
+ roughly speaking , follows from recent works on the height fluctuations of random monotone interfaces associated to dimer coverings of the hexagonal lattice @xcite , but it requires some work to really prove the precise statement we need .
the proof of proposition [ th : kenyon ] is given in detail in section [ app : kenyon ] .
+ once we have proposition [ th : kenyon ] , the proof of proceeds via a standard argument which we simply sketch : eq . , together with the fact that the mixing time of the dynamics in @xmath333 is @xmath352 , implies that the variation distance between @xmath227 and the law of @xmath340 is smaller than @xmath359 for all @xmath360 .
since the dynamics undergoes @xmath361 updates during the time interval @xmath362 $ ] , via a union bound and the equilibrium estimate one gets that the probability in is lower bounded by @xmath363 for @xmath2 large , which is the desired bound .
+ [ rem : cruciale ] it is important to notice that exactly the same proof gives , for some @xmath7 and for all @xmath154 , @xmath364\right)\ge 1-\frac c{l^{4 } } \end{aligned}\ ] ] ( just look at the equilibrium estimate in proposition [ th : kenyon ] ) .
* case b : * @xmath253 is such that @xmath365 .
+ here , a rather rough argument suffices .
remark first of all that the vertical coordinate @xmath366 of @xmath22 belongs to the interval @xmath367 $ ] .
call @xmath368 . by monotonicity , to show it is sufficient to prove that @xmath369 for a modified dynamics where only spins in @xmath370 evolve , starting from the `` @xmath3 '' configuration , with boundary conditions given by @xmath371 if @xmath372 and @xmath373 if @xmath374 .
note that @xmath370 is a discrete spherical cap , and that the boundary conditions just described are `` @xmath3 '' below its base , and `` @xmath0 '' elsewhere .
of course , @xmath4 in is understood to be the first time when all spins in @xmath370 are `` @xmath0 '' . + we note first of all that @xmath370 is contained in a parallelepiped @xmath249 whose base is a square of side @xmath375 and whose height is @xmath376 .
next , by monotonicity we see that @xmath4 is stochastically increased if we replace @xmath370 with @xmath249 , with `` @xmath3 '' boundary conditions below its base square , and `` @xmath0 '' everywhere else
. one can decompose @xmath249 into @xmath377 horizontal squares of side @xmath378 , stacked one on top of the other .
if @xmath377 were @xmath73 , we would simply have the evolution for @xmath37 of the two - dimensional ising model with `` @xmath0 '' boundary conditions in a square of side @xmath378 ( the `` @xmath0 '' boundary conditions on the top face of @xmath249 would compensate exactly the `` @xmath3 '' boundary conditions on the bottom face ) , and we know ( * ? ? ?
* theorem 1.3 ( a ) , case @xmath9 ) that in this case one has @xmath379 for some positive @xmath380 , if @xmath31 is large enough . via a standard monotonicity argument ( cf .
* proof of theorem 1.3 ( a ) , case @xmath381 ) for details ) this implies that , for the evolution in the parallelepiped @xmath249 , it is very unlikely that @xmath4 exceeds @xmath382 : @xmath383\le h\exp(-\gamma \ell)\end{aligned}\ ] ] if @xmath384 , which is actually stronger than the estimate we wished to prove .
* case c : * @xmath253 is such that @xmath385 \mbox{\;\;and\;\ ; } \phi\sin\theta\in\left[0 , \frac{(1+\delta)\sqrt 2}{\sqrt l}(\log l)^{3/4}\right].\end{aligned}\ ] ] + let @xmath386 be the point of @xmath283 with angular coordinates @xmath387 where @xmath388 will be chosen later , and let @xmath389 be a point of @xmath390 of minimal distance from @xmath386 .
+ for every @xmath391 , let @xmath392 be the spherical coordinates of the vector @xmath393 , i.e. , the spherical coordinates of @xmath56 with respect to the point @xmath394 .
+ [ th : cartesiano ] we have , for the point @xmath253 under consideration , @xmath395 + we omit the proof of lemma [ th : cartesiano ] , since it requires only tiresome but elementary computations : one first writes down the cartesian coordinates of @xmath389 and @xmath22 , then one works out the spherical coordinates of @xmath396 ; finally , the three statements are obtained by expanding these spherical coordinates for @xmath2 large , using the fact that @xmath397 .
+ we can now conclude the proof of , case * c*. first of all , remark that , thanks to , one has for @xmath2 large enough and letting @xmath398 , with @xmath399 the ball of radius @xmath242 centered at @xmath217 , @xmath400 provided that @xmath401 secondly , implies that for @xmath2 large enough @xmath402 provided that @xmath403 note that conditions and are compatible if @xmath274 is small enough , and choose a value for @xmath380 which satisfies both .
finally , shows that @xmath404 .
+ all in all , thanks to eq . in remark
[ rem : cruciale ] ( applied with @xmath2 replaced by @xmath405 , and modulo a trivial translation of the center of the sphere @xmath406 from @xmath407 to @xmath408 ) , we have @xmath409\right)\ge 1-c/(4l)^4\end{aligned}\ ] ] with @xmath135 the law of the dynamics in the sphere @xmath410 with `` @xmath0 '' boundary conditions on @xmath411 , started from `` @xmath3 '' . since @xmath412 , by monotonicity implies the same inequality when @xmath135 is the law of the dynamics in @xmath303 , always started from `` @xmath3 '' and with boundary `` @xmath0 '' conditions on @xmath413 .
the desired estimate is proven .
a subset @xmath415 is said to be _ monotone _
if @xmath416 implies @xmath200 whenever @xmath417 and @xmath418 .
the collection of all monotone sets is denoted by @xmath419 .
remark that this definition differs from that of _ positive monotone set _
definition [ def : monset ] ) only in that it is not required that @xmath198 . given a _
positive _ monotone set @xmath420 , in the following we will always implicitly identify it with the monotone set @xmath421 .
[ def : v ] given @xmath422 , we associate to it a _ vertical height function _ , which we denote @xmath423 .
the function @xmath424 is defined as @xmath425 observe that @xmath426 takes values in @xmath427 , and that @xmath428 we will denote @xmath429 the set of all possible functions @xmath430 which satisfy . as discussed in section [ sec : k ] , one can identify @xmath431 with @xmath432 , where @xmath433 is the set of dimer coverings of the infinite honeycomb lattice @xmath434 . recall
the definition of the monotone sets @xmath342 in section [ sec : ubmt ] , and note that the corresponding height functions @xmath435 coincide outside some domain @xmath436 whose diameter is @xmath437 ( cf .
lemma [ th : geomovvio](a ) ) .
note that @xmath438 is just the projection on the plane @xmath313 of the discrete spherical cap @xmath333 .
the estimate is proven if we have , for some universal constant @xmath7 , @xmath439 for @xmath154 where @xmath440 and @xmath441 ( for some @xmath442 ) denotes the uniform measure over the elements @xmath423 of @xmath431 such that @xmath443 for @xmath444 .
( since monotone sets and height functions are in one - to - one correspondence , we use the same notation @xmath445 to denote the equilibrium uniform measure in both cases ) .
note that the event @xmath262 is increasing with respect to the natural partial order in @xmath431 , where we say that @xmath446 ( with @xmath447 ) if @xmath448 for every @xmath449 . to lighten notations , given @xmath450
we will write @xmath451 if @xmath452 for every @xmath453 .
we need the following monotonicity property , which is an immediate consequence of proposition [ th : mondyn ] : [ th : monoton ] one has @xmath454 if @xmath455 are such that @xmath456 , @xmath457 and @xmath458 .
moreover , @xmath459 depends only on the value of @xmath460 on @xmath461 .
applying lemma [ th : monoton ] , we get @xmath462 the key point is the following result , whose proof is given in section [ sec : k ] : [ th : k ] there exists a probability measure @xmath463 on the elements @xmath464 which satisfies the following properties : @xmath465 where @xmath466 and @xmath31 is a universal positive constant .
we are now in a position to prove proposition [ th : kenyon ] .
we have @xmath467\nu(v^-_{\partial u}\le w_{\partial u})\\\nonumber & \ge & ( 1-\varepsilon_l ) \rho_{\left.v^-\right|_{u^{\tt c } } } ( a|\gamma^1)\end{aligned}\ ] ] where we used and lemma [ th : monoton ] .
therefore , @xmath468 where we used assumptions , , markov s inequality and the obvious bound @xmath469 . recalling the definition of @xmath470 , this implies .
@xmath471 the idea behind theorem [ th : k ] is that , while the height fluctuations of @xmath423 under @xmath445 for the _ fixed boundary conditions _ imposed by @xmath472 are hard to control , they are instead easily described if the boundary conditions are sampled from the infinite measure @xmath463 described in section [ sec : k ] .
such random boundary conditions are with high probability `` higher '' than the deterministic ones ( see ) and an application of monotonicity ( cf .
the steps in - ) concludes the argument .
we need first to recall some notions and results about the connection between monotone sets and dimer coverings of the infinite two - dimensional hexagonal lattice ( cf .
for instance @xcite ) . for @xmath275 , let @xmath473 denote its orthogonal projection on the @xmath474 plane @xmath475 .
let also @xmath476 and note that @xmath477 is a two - dimensional triangular lattice , which we consider as a graph by putting an edge between any two nearest neighbors ( whose mutual distance is @xmath478 ) .
we also let the vector @xmath479 ( resp . @xmath480 and @xmath481 )
be the @xmath482 projection of the vector which joins @xmath483 to @xmath484 ( resp .
to @xmath485 and to @xmath486 ) : of course , the @xmath487 have norm @xmath478
. see figure [ fig : hexa ] .
let @xmath434 be the dual lattice of @xmath477 , again with edges between nearest neighbors .
@xmath434 is a hexagonal lattice which we decompose as @xmath488 , where both @xmath489 and @xmath490 are translates of the triangular lattice @xmath477 , such that all nearest neighbors of every vertex in @xmath489 ( resp . in @xmath490 ) belong to @xmath490 ( resp . to @xmath489 ) .
vertices in @xmath489 ( resp . in @xmath490 ) are said to be _ white _ ( resp .
_ black _ ) .
* conventions * do not coincide with those of @xcite , but this is simply an irrelevant matter of convention .
] we embed @xmath477 and @xmath434 in @xmath491 , with the convention that @xmath492 is mapped to @xmath493 , that @xmath481 is mapped to the vertical vector @xmath494 and that @xmath479 is obtained by @xmath481 by a counter - clockwise rotation of @xmath495 .
also , given the two endpoints of the edge of @xmath434 which crosses @xmath481 , we decide that the one which has negative horizontal coordinate ( call it @xmath496 ) belongs to @xmath489 , while the other ( call it @xmath497 ) belongs to @xmath498 . observe that the vector @xmath499 is proportional to @xmath500 .
we will label a site @xmath501 ( resp . in @xmath490 ) as @xmath502 if @xmath503 ( resp . as @xmath504 if @xmath505 ) .
edges of @xmath434 which are perpendicular to @xmath479 ( resp . to @xmath480 or @xmath481 ) will be called edges _ of type `` @xmath217 '' _ ( resp . of type `` @xmath506 '' or `` @xmath31 '' ) .
see figure [ fig : hexa ] . and of the hexagonal lattice @xmath434 .
, scaledwidth=90.0% ] given a monotone set @xmath507 , we define the height function @xmath508 as follows : @xmath509 we denote by @xmath510 the set of all functions @xmath511 such that there exists @xmath422 with @xmath512 .
it is easy to see that there is a one - to - one mapping between elements of @xmath419 and elements of @xmath431 ( cf .
definition [ def : v ] ) , and between elements of @xmath431 and elements of @xmath510 .
in other words , the functions @xmath513 ( cf .
definition [ def : v ] ) and @xmath514 are two equivalent ways to describe the height of the set @xmath515 with respect to the horizontal plane .
a _ dimer covering _ @xmath516 of the hexagonal lattice @xmath434 is a subset of the edges of @xmath434 covering each vertex of @xmath517 exactly once .
note that each edge in @xmath516 covers one black and one white vertex .
to each height function ( or monotone set ) @xmath518 is uniquely associated a dimer covering , and conversely to a dimer covering one can associate uniquely a height function , provided that one fixes the height function at some arbitrary point @xmath519 ( in other words , dimer coverings identify only gradients of the height function ) .
is to associate to every @xmath416 a unit cube , centered at @xmath520 . in this way
, a monotone set , when seen from the @xmath521 direction , appears as a tiling of a portion of the plane with three types of lozenges ( top drawing on the left ) .
if we mark a segment along the longer diagonal of each lozenge ( top drawing on the right ) , we obtain a dimer covering of a portion of the hexagonal lattice ( bottom , left ) .
to each vertex of a lozenge , i.e. to every vertex in a triangular lattice , one associates its vertical height ( bottom , right ) .
one can check in the drawing that the dimer covering and the height function are linked by the construction explained in the text.,scaledwidth=90.0% ] the construction of the dimer covering @xmath516 given @xmath522 goes as follows ( see also figure [ fig:4 ] ) .
let @xmath523 be an edge of @xmath434 , and let @xmath313 be the edge of @xmath524 which intersects @xmath523 , with the convention that the vector @xmath525 is @xmath526 for some @xmath527 .
if @xmath528 or @xmath529 , then we put a dimer on the edge @xmath523 if @xmath530 , and we do not put it if @xmath531 ( it is easy to see that these are the only two possibilities since @xmath377 is a monotone interface ) .
if @xmath532 , then we put a dimer on @xmath523 if @xmath531 and we do not put it if @xmath533 .
note that the asymmetry between the indices @xmath534 is due to the fact that we are computing heights with respect to the horizontal plane . a more symmetric choice ( but less convenient for our purposes )
would be to measure heights with respect to the @xmath474 plane .
conversely , the construction of @xmath377 given a dimer covering @xmath516 goes as follows .
first we define a _ flux _
@xmath535 , i.e. a function on oriented edges @xmath523 of @xmath434 , such that @xmath536 .
if @xmath523 is oriented from the white to the black vertex , then : * if @xmath523 is of type `` @xmath217 '' or of type `` @xmath506 '' , then @xmath537 if @xmath538 and @xmath539 otherwise * if @xmath523 is of type `` @xmath31 '' , then @xmath537 if @xmath540 and @xmath541 otherwise . next , we fix some arbitrary value @xmath542 at some point @xmath543 .
finally , the difference @xmath544 for @xmath545 is the total flux of @xmath535 which crosses , from right to left , a path @xmath378 of edges of @xmath477 , which goes from @xmath389 to @xmath423 .
the fact that @xmath544 does not depend on the choice of the path @xmath378 is due to the fact that the flux @xmath535 has zero divergence , see ( * ? ? ?
2.2 ) . given @xmath546 with @xmath547 and @xmath548 ,
take a triangle of perimeter @xmath73 whose angles are @xmath549 and let @xmath550 ( resp .
@xmath551 ) be the length of the side opposite to @xmath552 ( resp .
@xmath553 ) . to every choice of @xmath554 as above , one can associate a translation - invariant gibbs measure @xmath555 on dimer coverings of @xmath434 .
translation - invariance means that , if @xmath262 is a set of edges of @xmath434 , then @xmath556 , where @xmath557 is a translation which maps @xmath434 into itself .
the precise statement is the following , whose different pieces were proved in @xcite : [ th : kos ] there exists a unique translation - invariant law @xmath555 on dimer coverings , such that the probability that a given edge of type `` @xmath217 '' ( resp . of type @xmath558 ) belongs to @xmath516 is @xmath559 ( resp .
@xmath560 ) and such that , conditionally on the configuration @xmath561 of the covering @xmath516 outside a given domain @xmath562 , @xmath555 is the uniform measure over all coverings @xmath563 of @xmath564 compatible with @xmath561 , i.e. , such that @xmath565 is a covering of @xmath434 ( we refer to this property as `` dlr property '' ) .
explicitly , @xmath555 is described as follows .
define the matrix @xmath566 as follows : * if @xmath567 is not a nearest neighbor of @xmath568 , then @xmath569 ; * @xmath570 ( resp .
@xmath551 ) if the edge @xmath571 ) is of type `` @xmath217 '' ( resp . of type @xmath558 ) .
define also the matrix @xmath572 as @xmath573 and @xmath574 where the integral is taken over the two - dimensional torus @xmath575 .
then , given a set of edges @xmath576 in @xmath434 , one has @xmath577
note that @xmath578 is a weighted version of the adjacency matrix of @xmath434 .
it is immediate to check that one has the relation @xmath579 , which justifies the notation @xmath580 .
the infinite matrix @xmath578 however does not admit a unique inverse , as discussed for instance in @xcite .
[ rem : pp ] if we fix deterministically the value @xmath581 for some @xmath543 , then the function @xmath582 is linear in @xmath583 , thanks to translation invariance of @xmath555 . as a consequence , the set of points @xmath584
is contained in some plane @xmath585 which of course contains the point @xmath586 such that @xmath587 and @xmath588 .
it is rather easy to check that the normal vector of the plane @xmath589 is parallel to @xmath590 . in particular ,
the plane @xmath589 is monotone in the sense of definition [ def : monotone ] .
set for lightness of notation @xmath591 .
[ th : flutt ] let @xmath543 and fix @xmath581 to some deterministic value .
there exists @xmath7 such that for every @xmath154 one has @xmath592 uniformly in @xmath590 ( where @xmath190 denotes the euclidean distance in @xmath477 ) .
_ proof of proposition [ th : flutt ] .
_ by translation invariance of @xmath555 , we can assume without loss of generality that @xmath593 and that we fixed @xmath594 . if @xmath595 , we can write @xmath596 for some @xmath597 . therefore , via a union bound , to show it is enough to prove that , for every @xmath598 @xmath599 for @xmath600 .
we consider for instance the case @xmath601 and @xmath259 , the other cases being essentially identical . from the construction of the height function in section [ sec : k ] , we know that @xmath602 i.e. , it is just the number of dimers in the set @xmath603 .
thanks to the determinantal representation , one has @xcite that @xmath604 has the same law as a sum of @xmath605 independent bernoulli random variables @xmath606 , whose parameters @xmath607 are the eigenvalues of the matrix @xmath608 . in general , it is not easy to solve explicitly the double integral which defines @xmath580 .
however , there are special values of @xmath609 for which @xmath610 takes an easy form .
in particular , one checks that , probably due to a typo there . in any case , the global sign is inessential for our computation , since below depends only on the absolute value of @xmath580 .
] , for @xmath611 , @xmath612 and that @xmath613 as a consequence , the entries of the matrix @xmath262 are given by @xmath614}{\pi(i - j ) } & i\ne j\\ p_a&i = j \end{array } \right.\end{aligned}\ ] ] and then @xmath615 we will show in a moment that @xmath616 for some @xmath617 , uniformly in @xmath590 , which allows to conclude the proof of : via the exponential tchebyshev inequality , @xmath618 using the inequality @xmath619 which holds for @xmath620 and the fact that @xmath621 , one deduces @xmath622 analogously , one obtains the same upper bound for @xmath623 and is proven .
it remains only to prove the estimate .
essentially the proof can be found in ( * ? ? ?
6.3 ) , where however uniformity with respect to @xmath590 was not discussed , so we repeat quickly the necessary steps here .
observe first of all that @xmath624 uniformly in @xmath590 .
therefore , @xmath625 + 2\sum_{i=1}^{n-1}i\,a_i^2 + 2n\sum_{i = n}^\infty a_i^2\\ & \le & n\left[a_0(1-a_0)-2\sum_{i=1}^\infty a_i^2 \right]+c\log n\end{aligned}\ ] ] for some @xmath31 independent of @xmath590 .
finally , one observes that the quantity in square brackets is identically equal to zero . to see this ,
let @xmath626 so that @xmath627 .
one then observes that @xmath628 and then a straightforward computation shows that the fourier identity @xmath629 is equivalent to @xmath630 .
@xmath631 _ proof of theorem [ th : k ] .
_ recall the setting of section [ sec : eq17 ] .
let @xmath632 , let @xmath633 , and @xmath634 .
choose @xmath590 to be parallel to the normal to the plane @xmath293 of definition [ def : geom](c ) and consider the infinite volume distribution of height functions induced by @xmath555 , with the normalization @xmath635 .
note that the plane @xmath636 defined in remark [ rem : pp ] is obtained from @xmath358 by translating it for a distance @xmath637 in the positive vertical direction , so that @xmath638 . for a given realization of the height function @xmath377 , we let @xmath639 be the finite set of points @xmath640 thanks to proposition [ th : flutt ]
, one has that @xmath641 with @xmath31 independent of @xmath554 . recall that to a height function @xmath518 there corresponds a unique element @xmath642 and let @xmath463 be the law on @xmath431 induced by @xmath643 .
note that @xmath644 deterministically .
we show now that @xmath463 satisfies the conditions . the point is that , as observed at the beginning of section [ sec : eq17 ] , the diameter of @xmath645 is @xmath646 , so that the @xmath482 projection of any point @xmath647 , with @xmath648 , has distance from @xmath389 smaller than @xmath2 . as a consequence , given some @xmath649 , @xmath650 and , from , @xmath651 which immediately implies condition ,
since the distance between @xmath589 and @xmath358 is @xmath652 , and the graph of the function @xmath653 is within distance @xmath654 from the plane @xmath358 ( recall that @xmath655 is the vertical height function of the monotone set @xmath656 defined in so that , in @xmath438 , @xmath655 is just the lattice approximation of the plane @xmath358 ) .
conditions , are also immediate from , once one realizes that @xmath657 if @xmath658 is an event which depends only on @xmath659 .
this is because , thanks to the dlr property of @xmath643 ( cf .
theorem [ th : kos ] ) , the measure @xmath660 is the uniform measure over the elements of @xmath431 which coincide with @xmath423 outside @xmath438 ; in other words , @xmath660 is nothing but @xmath661 .
in this section we prove theorem [ th : wilson ] . recall the notation from section [ sec : wilson ] , in particular the definition of @xmath663 , of @xmath664 ( the diameter of the horizontal projection of @xmath16 ) and of @xmath238 ( the maximal vertical distance between two points in @xmath16 with the same horizontal projection ) . without loss of generality
, we can assume that @xmath16 is contained in some parallelepiped @xmath665 defined by @xmath666 with @xmath667 and @xmath668 , @xmath669 .
recalling definition [ def : v ] , one can identify uniquely any monotone set @xmath420 , such that @xmath670 , via a vertical height function @xmath671 , indexed by pairs @xmath313 of non - negative integers . since the configuration of @xmath515
outside @xmath16 is fixed and coincides with @xmath672 , what matters is the collection @xmath673 .
this defines a _ plane partition _ in the box @xmath674 ( cf .
also definition [ def : monset ] ) , i.e. a collection of heights such that @xmath675 , @xmath676 . as discussed in section [ app : kenyon ] , the bijective correspondence between @xmath423 and @xmath515 is given by @xmath677 each pair @xmath313 can be identified with a unit square in the plane @xmath678 , in such a way that the center of this square is given by the point of @xmath290 with coordinates @xmath679 .
thus , we interpret @xmath671 as the height of the column at @xmath313 .
the minimal and maximal sets @xmath680 correspond to minimal and maximal column heights , denoted @xmath655 and @xmath681 respectively ( the same notation was used in section [ sec : eq17 ] ) .
we define @xmath682 as the set of all plane partitions @xmath423 in the box @xmath674 such that @xmath683 .
thus , @xmath684 is in one - to - one correspondence with the set @xmath685 such that @xmath686 , and the measure @xmath687 of section [ sec : wilson ] becomes now the uniform probability measure on @xmath684 ( we still call it @xmath688 ) .
the first step in the proof of theorem [ th : wilson ] consists in establishing a mixing time upper bound ( lemma [ le : tmixbo ] below ) for a markov chain that involves equilibration of full columns at each move .
the key idea here borrows from wilson s analysis @xcite of the luby - randall - sinclair markov chain for lozenge tilings @xcite .
the second step ( see section [ sec : compare ] ) is to show that lemma [ le : tmixbo ] implies the upper bound on the mixing time of the `` single spin - flip '' dynamics @xmath689 which is under consideration in theorem [ th : wilson ] .
a similar strategy was used in @xcite in the simpler context of the @xmath690-dimensional sos model .
a column @xmath313 is said to be even / odd if @xmath691 is even / odd .
consider the continuous time markov chain with state space @xmath684 , where at each arrival time of a poisson process with parameter @xmath73 we flip a fair binary coin ; if the coin is @xmath407 ( resp .
@xmath73 ) we update simultaneously all even ( resp .
odd ) column heights @xmath671 with a sample from the distribution @xmath688 conditioned on the current value of the height of the odd ( resp .
even ) columns .
let @xmath692 denote the distribution at time @xmath24 of such markov process when the starting configuration is @xmath693 .
note that the kernel @xmath694 satisfies @xmath695 where the infinitesimal generator @xmath696 acts on functions @xmath697 by @xmath698(v ) = \frac12\ , { \rho}[\,f{\ , | \,}{\rm odd}\ , ] ( v ) + \frac12\ , { \rho}[\,f{\ , | \,}{\rm even}\,](v ) - f(v)\,,\ ] ] where we use the notation @xmath699 ( v)$ ] ( resp . @xmath700 ( v)$ ] ) for the expectation with respect to @xmath688 conditioned on @xmath701 ( resp .
@xmath702 ) . note that sampling from @xmath703 $ ] amounts to pick uniformly at random a new configuration of even column heights that is compatible with the current odd column heights ( compatibility here means that the configuration of all column heights is then a plane partition @xmath423 that satisfies @xmath704 ) .
it is important to remark that the probability measure @xmath703 $ ] is a product of single column probability measures , i.e. conditionally on @xmath701 , all even columns become independent .
the same remarks apply to @xmath705 $ ] .
clearly , @xmath688 is the reversible invariant distribution of our markov chain .
[ le : tmixbo ] if @xmath234 , there exists a constant @xmath7 such that @xmath706 we use the well known `` lattice path representation '' of a plane partition ; see e.g. @xcite .
namely , any plane partition @xmath693 can be seen as a collection of @xmath707 non - intersecting lattice paths @xmath708 , @xmath709 , each of length @xmath710 ( @xmath711 are the same integers that appear in ) which satisfy @xmath712 for all @xmath713 .
the polymer @xmath714 describes the @xmath715 level set @xmath716 . for the precise construction of the paths
, we refer to ( * ? ? ?
* section 5 ) , ( * ? ? ?
* section 2.1 ) ( see also figure [ fig : polymers ] for a graphical construction ) . .
in the right drawing , the corresponding `` lattice path '' representation . the lattice path @xmath717 is essentially the level set at height @xmath714 of the column representation of @xmath515 , seen from the @xmath474 direction .
it is clear that , whenever we _ add _ a cube to the plane partition on the left , one lattice path on the right will be _
lowered_. this shows that @xmath718 is reversed into @xmath719.,scaledwidth=80.0% ] in the plane partition - to - lattice path mapping , inequalities are reversed , i.e. if @xmath718 then @xmath719 ( cf . figure [ fig : polymers ] ) .
we let @xmath720 and @xmath721 denote the lattice paths corresponding to the maximal and minimal plane partitions @xmath722 , so that @xmath723
. then , the condition @xmath704 gives @xmath724 for all @xmath713 or , more compactly , @xmath725 .
let @xmath726 denote the set of configurations @xmath727 of integer heights @xmath728 , @xmath729 and @xmath730 satisfying the constraints , . the construction above establishes a one - to - one correspondence between the set @xmath684 of plane partitions @xmath423 satisfying @xmath683 and the set @xmath726 . the image of the measure @xmath688 is the uniform probability distribution on @xmath726 ( which we call again @xmath688 with some abuse of language ) .
moreover , it is not hard to see that the image of the markov process with `` full column moves '' under this map coincides with the continuous time markov process with state space @xmath726 obtained as follows : at each arrival time of a poisson process with parameter @xmath73 we flip a fair binary coin ; if the coin is @xmath407 ( resp .
@xmath73 ) we update simultaneously all @xmath731 , @xmath709 , for @xmath22 even ( resp .
odd ) with a sample from the uniform distribution conditioned on the current values @xmath732 ( resp .
even ) . with a slight abuse of notation
we call again @xmath733 the kernel of the markov process on lattice path configurations , and its invariant measure is of course the uniform measure @xmath688 . moreover , we write again ( cf . ) @xmath734 - \mathbb
i \big ) + \frac12\,\big({\rho}[\,\cdot{\ , | \,}{\rm even}\ , ] - \mathbb i \big)\,,\ ] ] for the generator of @xmath735 , where @xmath736 denotes the identity operator .
next , we turn to wilson s coupling argument
. define @xmath737 by @xmath738 to compute the action of @xmath696 on @xmath739 , fix some even @xmath740 , and observe that if the constraints @xmath741 were absent , one would have @xmath742 = \frac12\sum_{j = h_1}^{h_2 } ( \phi^{(j)}_{x-1}+ \phi^{(j)}_{x+1})\,.\ ] ] note that the sum is important in since the identity has no reason to hold for a single @xmath714 .
now , the constraint is felt at @xmath22 in the path @xmath714 iff either @xmath743 or @xmath744 . in the first case we have to compensate with @xmath745 since @xmath746 can not move and @xmath747 . in the second case
we have to compensate with @xmath39 since @xmath746 can not move and @xmath748 . in conclusion ,
@xmath749 = \frac12\sum_{j = h_1}^{h_2 } ( \phi^{(j)}_{x-1}+ \phi^{(j)}_{x+1 } ) + \sum_{j = h_1}^{h_2 } ( 1_{a_+(j , x)}(\phi)-1_{a_-(j , x)}(\phi))\,.\ ] ] clearly , for any @xmath22 even one has @xmath750=
\sum_{j = h_1}^{h_2 } \phi^{(j)}_x$ ] .
moreover , exactly the same equations hold when @xmath22 is odd , provided we change the conditioning from odd to even . therefore , using the notation @xmath751 for the discrete laplacian , together with its analogue for odd @xmath22 imply that @xmath752(\phi ) = \frac12 \sum_{x=1}^{2d-1 } g(x)\sum_{j = h_1}^{h_2 } ( \delta\phi^{(j)})(x ) + \sum_{x=1}^{2d-1 } g(x)\sum_{j = h_1}^{h_2 } ( 1_{a_+(j , x)}(\phi)-1_{a_-(j , x)}(\phi)).\ ] ] summing by parts and using @xmath753 , where @xmath754 , one has @xmath755(\phi ) = -\frac{\kappa_d}2\ , \phi(\phi)+ \sum_{x=1}^{2d-1 } g(x)\sum_{j = h_1}^{h_2 } ( 1_{a_+(j , x)}(\phi)-1_{a_-(j , x)}(\phi)).\ ] ] next , let @xmath756 denote the state of the markov process at time @xmath24 with initial condition @xmath757 at time @xmath407 .
when @xmath758 ( minimal state in terms of lattice paths ) or @xmath759 ( maximal state ) , we simply write @xmath760 or @xmath761 . define @xmath762\ge0 $ ] , where @xmath109 denotes expectation with respect to the global monotone coupling of the lattice path markov process with kernel @xmath735 ( cf .
section [ sec : monotone ] and @xcite ) . from we
infer @xmath763 with @xmath764\,.\ ] ] by monotonicity of the coupling it is immediate to see that @xmath765 .
therefore implies @xmath766 . note that @xmath767 can be upper bounded by the volume enclosed between the minimal and maximal plane partition , i.e.@xmath768 ( recall that in theorem [ th : wilson ] we are assuming @xmath234 )
it follows that @xmath769 .
to finish the proof it suffices to observe that by monotonicity @xmath770 and @xmath771 iff @xmath772 , so that by markov s inequality @xmath773 we can now bound the total variation distance by the probability of no coupling up to time @xmath24 , and using monotonicity and the bound above this gives , for any initial state @xmath423 @xmath774 which is easily seen to imply the desired estimate .
@xmath775 the proof of theorem [ th : wilson ] uses the bound of lemma [ le : tmixbo ] together with a comparison argument that allows us to translate the mixing time upper bound of the `` column dynamics '' to an upper bound for the mixing time of the dynamics @xmath689 defined in section [ sec : wilson ] . in the plane partition language , the dynamics @xmath776 will be called @xmath777 , with law @xmath778 and equilibrium distribution as usual denoted by @xmath688 : it has initial condition @xmath779 ( @xmath780 being the plane partition corresponding to the monotone set @xmath781 ) , to each column @xmath313 is associated an independent poisson clock with parameter @xmath73 and the evolution proceeds by local updates . by this
we mean that when the clock of the column @xmath313 rings , one replaces @xmath671 by @xmath782 ( with probability @xmath218 ) or by @xmath783 ( with probability @xmath218 ) .
in other words , the chosen column performs a simple symmetric random walk step , except that jumps which violate the plane partition constraints @xmath784 , @xmath785 , or the overall constraint @xmath786 , are rejected .
we need to prove that the mixing time of this chain is @xmath787 .
we start with a simple observation that allows us to reduce to the case of maximal ( @xmath788 ) and minimal ( @xmath789 ) initial conditions . [
le : monopm ] for any @xmath790 and any @xmath791 : @xmath792
let @xmath108 denote a monotone coupling of @xmath793 . then , using the fact that each column @xmath313 has a minimal height @xmath794 and a maximal height @xmath795 such that @xmath796 we have @xmath797 \leq d^3\,\|\nu_t^{v^+ } - \nu_t^{v^-}\|\,.\end{aligned}\ ] ] @xmath798 thanks to lemma [ le : monopm ] , to prove theorem [ th : wilson ] it is sufficient to show that @xmath799 for some @xmath800 .
let us prove the statement for @xmath801 , the argument for @xmath802 being identical .
consider i.i.d .
bernoulli(@xmath803 ) random variables @xmath804 and let @xmath805 , with @xmath806 and @xmath807 , denote a partition of the time axis .
furthermore , consider i.i.d .
poisson processes with parameter @xmath73 at each column @xmath313 , and let @xmath808 denote the corresponding collection of arrival times . call @xmath809 the collection of arrival times obtained from @xmath808 by deleting ( or `` censoring '' ) all arrivals @xmath810 such that , for some @xmath811 , @xmath812 and @xmath313 has the opposite parity as @xmath813 ( e.g. @xmath313 is odd and @xmath814 ) . by construction , if we start from the configuration @xmath681 at time @xmath407 and perform local updates using all the marks @xmath808 up to time @xmath24 we obtain the distribution @xmath815 .
let us call @xmath816 the distribution of the `` censored '' dynamics obtained in the same way but only using the marks @xmath817 , for a fixed bernoulli sequence @xmath818 .
from the `` censoring inequality '' of peres and winkler @xcite , ( * ? ? ? * th .
2.5 ) it follows that @xmath801 is stochastically dominated by @xmath819 , for any @xmath820 . moreover , setting @xmath821 , where @xmath822 denotes expectation over the random sequence @xmath818 , by linearity of the expectation one sees that @xmath823 let us fix @xmath824 , where @xmath825 and @xmath826 are constants to be taken sufficiently large .
let @xmath827 denote a poisson process with parameter @xmath73 , and write @xmath108 and @xmath109 for the associated probability and expectation .
monotonicity implies that the event @xmath828 is increasing ( * ? ? ?
therefore , the censoring inequality gives @xmath829 taking @xmath830 large in the definition of @xmath831 above and using standard estimates for the poisson random variable we can make @xmath832 smaller than @xmath833 .
therefore , we see that thanks to lemma [ le : tmixbo ] and , if @xmath834 in the definition of @xmath74 is chosen large the claim is a consequence of @xmath835 where @xmath836 is the kernel defined in section [ sec : columns ] , which involves full column equilibrations ( cf . ) . to prove we need to compare column equilibration moves with local updates .
let us use the notation @xmath837 and @xmath838 for the probability kernels associated to @xmath839(v)$ ] and @xmath840(v)$ ] respectively , see .
that is , for any @xmath697 one has for instance @xmath841(v)=\sum_{v'\in{\omega}}{\rho}_0(v , v')f(v')$ ] .
define @xmath842(v,\cdot)$ ] , and note that @xmath843 is nothing else but a discrete time version of the kernel @xmath844 , i.e. @xmath845 where @xmath109 and @xmath846 are as above . since @xmath825 , excluding an event of probability @xmath847 for some large @xmath848 we can assume that @xmath849 for some @xmath850 .
recall that @xmath816 denotes the law @xmath851 conditioned on the binary sequence @xmath818 .
the previous remarks imply that it will be sufficient to prove the upper bound @xmath852 observe that , for the censored dynamics @xmath853 , in the time interval @xmath854 , all columns with the same parity as @xmath855 are independently updated by local moves , i.e. on columns of that parity we have independent continuous - time simple symmetric random walks in segments ( determined by the columns of opposite parity ) of length bounded by @xmath238 .
it is standard that for some constant @xmath7 the mixing time on each column is bounded by @xmath856 and therefore after a time @xmath557 we have the bound , for any @xmath857 , uniformly in the starting configuration @xmath693 : @xmath858 for general @xmath859 , by recursive coupling of the distributions involved , using at each step , one has @xmath860(v^+,\cdot ) \| \leq 1-\big(1-c\,d^2\,\exp{\big(-\frac{t}{c\,h^2}\big)}\big)^n\,.\ ] ] since @xmath806 and @xmath861 , we see that the desired estimate follows for a suitable choice of @xmath862 .
here we prove the bound @xmath864 for a suitable constant @xmath31 .
let @xmath865 be the cube @xmath866 ( we assume for definiteness that @xmath2 is even ) with boundary conditions @xmath113 if @xmath867 with @xmath868 and @xmath869 otherwise .
in other words , the boundary conditions @xmath92 are `` @xmath0 '' at the three faces of @xmath870 which meet at the origin of @xmath390 , and `` @xmath3 '' at the other three . denote by @xmath871 the zero - temperature glauber evolution started from the `` @xmath3 '' configuration ( in order not to confuse it with the evolution @xmath340 in @xmath250 ) and by @xmath872 its law .
one has [ prop : spigolo ] there exists @xmath7 such that @xmath873 _ proof of eq .
, assuming proposition [ prop : spigolo]_. since the set @xmath874 is a monotone set ( cf . definition [ def : monset ] ) at all times , the event that @xmath875 implies the event @xmath876 for all @xmath877 such that @xmath878
. therefore , implies ( using also symmetry among the three coordinate axes ) @xmath879 the cube @xmath880 can be seen as the union of eight disjoint sub - cubes @xmath881 of side @xmath882 , with @xmath883 while @xmath884 are suitable translations of @xmath865 .
let as usual @xmath135 denote the law of the glauber evolution inside @xmath250 , with `` @xmath0 '' b.c .
, started from `` @xmath3 '' and let @xmath885 be law of the evolution , again started from `` @xmath3 '' , where the spin configuration inside each @xmath881 evolves independently for different @xmath886 , with b.c . given by @xmath113 for @xmath887 and @xmath869 for @xmath888 .
it is clear that , until the random time @xmath889 the two evolutions can be perfectly coupled , and that @xmath890 . on the other hand , thanks to and to the symmetry among the various cubes @xmath881 ( up to suitable translations of the origin and reflections of the coordinate axes ) one sees that @xmath891 @xmath892 the main ingredient in the proof of proposition [ prop : spigolo ] is a `` column dynamics '' @xmath893 ( analogous to the one used in section [ sec : columns ] ) defined as follows .
start from the `` all minus '' configuration in @xmath865 , @xmath894 .
assign to each @xmath895 an i.i.d .
poisson clock of rate @xmath896 ; when the clock labeled @xmath897 rings , we assign a new value to the collection of spins at sites @xmath898 with @xmath899 , by sampling it from the equilibrium distribution conditioned on the present value of all the other spins . in other words
, we set to equilibrium the column of horizontal coordinates @xmath900 , conditionally on the value of the neighboring columns . for every @xmath24 we have the stochastic domination @xmath901 indeed , for @xmath902 let @xmath903 be the following dynamics .
set @xmath904 and , when the poisson clock of the column labeled @xmath22 rings , repeat @xmath905 times the following procedure : * flip a fair binary coin ; * if the coin gives `` head '' , then make a heat bath update at each of the sites of the column @xmath22 with vertical coordinate belonging to @xmath906 , one by one , starting from the bottom site ; * if instead the coin gives `` tail '' , then make a heat bath update at each of the sites of the column @xmath22 with vertical coordinate belonging to @xmath907 , one by one , starting from the bottom site . here , ``
making a heat - bath update '' at a site @xmath212 means updating @xmath908 according to the equilibrium conditioned on the value of the spins outside @xmath212 .
it is immediate to realize that the process @xmath909 has the same law as @xmath871 , and that the law of @xmath910 converges to that of @xmath911 for @xmath912 .
also , the stochastic domination @xmath913 is an immediate consequence of the peres - winkler censoring inequality @xcite , ( * ? ? ?
call @xmath914 the law of the column dynamics @xmath911 .
one has [ prop : coldyn ] fix @xmath7 . for @xmath2 sufficiently large and @xmath915 one has @xmath916 for some finite constant @xmath917 .
_ proof of proposition [ prop : spigolo ] , given proposition [ prop : coldyn]_. let @xmath918 so that , thanks to proposition [ prop : coldyn ] and to the stochastic domination one has @xmath919 note that the desired bound can be rewritten as @xmath920 . one has @xmath921 where in the first term we applied markov s inequality .
choosing @xmath31 sufficiently large , one can make both terms in the last expression smaller than @xmath11 .
indeed , in order that @xmath922 , there must be two times @xmath923 with @xmath924 such that the poisson clock associated to the site @xmath925 rings both at times @xmath74 and @xmath24 . via a simple union bound , and using the exponential form of the law of the intervals between two successive rings , one sees that the probability of such event is @xmath926 .
@xmath927 _ proof of proposition [ prop : coldyn]_. since the set @xmath928 is a monotone subset of @xmath865 at all times , one can identify it ( recall section [ sec : columns ] and fig .
[ fig : polymers ] ) with the set of @xmath882 paths @xmath929 of length @xmath930 , where the @xmath715 path is the collection @xmath931 and the following relations are satisfied : @xmath932 at time @xmath25 one has @xmath933 and therefore @xmath934 if @xmath935 and @xmath936 if @xmath937 . defining @xmath938 , \end{aligned}\ ] ] @xmath939 and reasoning like in section [ sec : columns ] ( see also ( * ? ? ?
5 ) ) , one sees that @xmath940 satisfies the discrete heat equation @xmath941 with dirichlet boundary conditions @xmath942 . on the other hand , from the representation of monotone sets as collections of paths
, one sees that @xmath943 therefore , it is sufficient to prove the heat - equation estimate @xmath944 uniformly for @xmath945 and @xmath2 large .
while can be obtained directly via fourier analysis , we give a simple and more probabilistic argument .
first it is well known ( cf .
for instance @xcite ) that , for @xmath946 , the quantity @xmath947/2 $ ] coincides with the probability that there is a particle at site @xmath22 at time @xmath24 , for a symmetric simple exclusion process on @xmath948 with initial condition at time zero such that sites @xmath935 are occupied by a particle , while sites @xmath949 are empty ( each particle attempts with rate one to jump to one of its two neighboring sites with equal probability @xmath218 and the jump is rejected if either the site is already occupied or if it lies outside @xmath948 ) .
in particular ( recall that @xmath950 ) one has that @xmath951/2 $ ] is the probability that there is a particle at @xmath2 at time @xmath24 . by duality ( cf .
* section ii.3 ) ) , this equals the probability that a continuous - time simple random walk of rate @xmath73 on @xmath948 , started from site @xmath2 , is in @xmath952 at time @xmath24 .
the bound then follows from standard random walk estimates : if @xmath953 is the law of the continuous - time simple random walk @xmath954 on @xmath955 started from @xmath22 , one has @xmath956 and the latter expression is easily seen ( e.g. using the local central limit theorem ) to be upper bounded by the r.h.s . of .
as discussed in remark [ rem : d2tmix ] , we only have to prove that , for the @xmath37 dynamics , @xmath958 for suitable positive constants @xmath959 .
also , thanks to monotonicity , it is enough to prove this fact for the dynamics @xmath340 in the domain @xmath960 with `` @xmath0 '' boundary conditions on @xmath961 .
the advantage of looking at the dynamics in @xmath962 instead of @xmath250 will be apparent in the proof of theorem [ th : restogood ] .
we need a certain number of geometric definitions : [ def : gset ] given @xmath963 , let a. @xmath964 b. @xmath965 where @xmath966 is the unit square of side @xmath73 centered at @xmath22 , with sides parallel to the coordinate axes .
the boundary of @xmath967 is denoted as @xmath968 and its geometric length as @xmath969 ; c. for @xmath600 @xmath970 and similarly @xmath971 d. @xmath972 be the configuration in @xmath973 obtained by flipping every `` @xmath3 '' spin in @xmath962 such that a strict majority of its nearest neighbors are `` @xmath0 '' , and repeating the same operation as long as such a site exists .
we call @xmath974 the `` majority transformation '' ; e. @xmath975 ; f. @xmath976 be the subset of @xmath977 defined by @xmath978 note that the constraint @xmath979 is simply the requirement that @xmath29 minimizes the hamiltonian @xmath980 , given the values @xmath981 .
next we introduce a modified dynamics @xmath982 on @xmath983 , with initial condition @xmath984 , as follows .
let @xmath985 and set @xmath986 for @xmath987 .
whenever the poisson clock labeled @xmath22 rings at a time @xmath988 , first refresh the current value of the spin at @xmath22 according to the distribution @xmath989 , and then apply the majority transformation @xmath974 of definition [ def : gset](d ) to the configuration thus obtained .
we call @xmath990 the generator of such dynamics and @xmath991 its law at time @xmath24 . when the initial condition is @xmath992 , it is immediate to see that , by the monotonicity of the usual glauber dynamics , one has @xmath993 for every @xmath790 , so that is proven if we show that @xmath994 for some suitable @xmath995 , where @xmath914 is the law of the process @xmath996 .
the advantage of the modified dynamics @xmath997 with respect to the usual glauber dynamics is that it belongs to the `` good set '' @xmath998 for all times : [ th : restogood ] for every @xmath790 one has that @xmath999 .
moreover , there exists a deterministic positive constant @xmath1000 such that @xmath1001 _ proof of theorem [ th : restogood ] .
_ for this , we need some additional notions : [ def : mvw ] given @xmath1002 and @xmath1003 , we say that @xmath22 is a `` flippable site '' if two neighbors of @xmath22 , at mutual distance @xmath1004 , are `` @xmath0 '' and the other two neighbors are `` @xmath3 '' . if @xmath22 is flippable , recall that @xmath1005 is the configuration obtained by flipping @xmath20 to @xmath1006 . then , for @xmath22 flippable we say that * @xmath22 is a `` mountain '' if @xmath1007 and @xmath1008 ; * @xmath22 is a `` vertex '' if @xmath1007 and @xmath1009 ; * @xmath22 is a `` valley '' if @xmath1010 .
[ rem : vertici ] observe that , if @xmath22 is a `` vertex '' , then exactly one of its neighbors , call it @xmath56 , is also a vertex .
it is immediate to see that @xmath1011 , cf .
figure [ fig : dl ] . , the sub - domain @xmath664 ( enclosed by the dashed line ; proportions are not respected ) and the contour @xmath967 ( thick line ) of a configuration @xmath1012 .
big full dots denote `` vertex '' sites ( remark that they occur in nearest - neighboring pairs ) , empty dots denote `` mountains '' and @xmath1013 denote `` valleys '' .
vertices can occur only in the right - most or left - most column which intersects @xmath1014 , or in the highest or lowest row which intersects @xmath1014 , so there are plainly at most @xmath1015 vertices .
the curve @xmath967 can be seen as the union of four monotone curves , which in the figure are delimited by the dotted lines . for each monotone curve
, one has @xmath1016.,scaledwidth=60.0% ] we prove that @xmath1017 for @xmath1018 by induction : clearly the statement is true at time zero , and we show that if it is true until some time @xmath74 then it remains true after the next update .
the first observation is that , if @xmath1019 , then nothing happens in the evolution until the poisson clock of a _
flippable site @xmath22 _ rings ( the occurrence of sites with three neighbors of opposite sign , or sites with exactly two neighbors of opposite sign at mutual distance @xmath896 is forbidden by the condition @xmath1020 )
. when such a ring happens , we have three possibilities : 1 .
@xmath22 is a valley .
then , with probability @xmath218 the configuration remains unchanged , and with probability @xmath218 we change @xmath1021 to @xmath1022 , cf .
definition [ def : mvw ] .
we do not need to apply the transformation @xmath974 , since one sees easily that @xmath1023 .
also , it is easy to see that @xmath1024 , and that @xmath1025 remains a simple curve , see figure [ fig : mvw ] .
2 . @xmath22 is a vertex .
with probability @xmath218 the configuration remains unchanged , and with probability @xmath218 we change @xmath1021 to @xmath1022 ; then , the application of the transformation @xmath974 has the effect of flipping also the vertex neighbor of @xmath22 , cf .
remark [ rem : vertici ] .
altogether , @xmath1026 remains a simple curve ; its length decreases by @xmath896 , but also does the sum @xmath1027 , see figure [ fig : mvw ] .
3 . @xmath22 is a mountain .
this case is more subtle , since it is not obvious apriori that @xmath1025 is a simple curve .
just to fix ideas , assume that the `` @xmath0 '' neighbors of @xmath22 in @xmath1021 are @xmath1028 . if @xmath1025 were not a simple curve , it would mean that @xmath1029 . since @xmath1020 , one has that the set @xmath1030 belongs to @xmath1031 ( so in particular it has no intersection with @xmath664 ) , otherwise the condition @xmath1032 would be violated .
however , by the definition of the domains @xmath962 and @xmath664 , there exists no site @xmath1033 such that @xmath1034 . indeed , for that to happen one would need that @xmath897 with either @xmath1035 and @xmath1036 or @xmath1037 and @xmath1038 , which is clearly incompatible with @xmath1003 , cf . .
this shows that @xmath1025 is a simple curve .
observe that it is for this issue that it was important to change the shape of the domain from @xmath1039 to @xmath983 .
of course , the value @xmath1040 in the definition of @xmath664 could be changed to any number larger than @xmath218 . in all cases ,
the configuration belongs to @xmath998 after the move , and the proof of @xmath1041 for all @xmath1042 is complete .
( @xmath22 is marked by a dot ) and of its neighboring vertex site from the value `` @xmath3 '' to `` @xmath0 '' . in this case , @xmath1043 decreases by @xmath73 ( while @xmath1044 and @xmath1045 are unchanged ) and @xmath967 remains a simple curve , but its length decreases by @xmath896 . in the middle drawing , the effect of flipping from `` @xmath3 '' to `` @xmath0 '' a `` mountain '' @xmath20 .
@xmath967 remains a simple curve and neither its length nor the values @xmath1046 vary .
similarly , the third drawing shows the effect of flipping a `` valley '' spin . in the three cases , only a portion of @xmath1047 is drawn.,scaledwidth=60.0% ] to prove ,
assume that the last update before @xmath1048 consisted in flipping from `` @xmath3 '' to `` @xmath0 '' a `` mountain '' site @xmath1049 ( if instead the move consisted in flipping from `` @xmath3 '' to `` @xmath0 '' a `` vertex '' site , the proof which follows would be very similar ) .
call @xmath1021 the spin configuration just before the last update and , just to fix ideas , assume that the two `` @xmath3 '' neighbors of @xmath22 in @xmath1021 are @xmath1050 and @xmath1051 . since @xmath22 is a mountain and @xmath1020 ,
it is immediate to realize that the set @xmath1052 is a subset of @xmath1053 , otherwise the condition @xmath1032 would be violated .
next , from the definition of the set @xmath664 one sees that @xmath1054 has cardinality at least @xmath1055 , uniformly in @xmath1056 ( explicitly , one can take @xmath1000 to be slightly less than @xmath1057 ) . the estimate is then proven .
@xmath1058 the simplification of considering a dynamics which evolves in the set @xmath696 is that for such configurations the numbers of valleys , mountains and vertices satisfy simple relations : [ th : diamante ] given @xmath1002 , let @xmath1059 ( resp .
@xmath1060 ) be the number of valleys ( resp .
mountains , vertices ) in @xmath962 . then , @xmath1061 and @xmath1062 .
the proof of lemma [ th : diamante ] is best explained through a picture , and therefore we refer to the caption of figure [ fig : dl ] .
now we can finish the proof of .
thanks to theorem [ th : restogood ] , we see that follows if we have @xmath1063 by the exponential tchebyshev inequality one has for @xmath1064 @xmath1065\\ & = & e^{-\lambda\psi l^2}\phi(c_0 l^2)\end{aligned}\ ] ] where @xmath1066 $ ] and @xmath1067 note that one has @xmath1068 and @xmath1069 where we recall that @xmath1070 is the law of @xmath997 and @xmath1071 is its generator . by the very definition of the modified dynamics , if @xmath1072 then @xmath1073 vanishes for every @xmath1074 . if instead @xmath1075 , we know from theorem [ th : restogood ] that we need only to consider the case @xmath1012 .
therefore , for @xmath1075 one has ( recall the discussion after remark [ rem : vertici ] ) @xmath1076.\end{aligned}\ ] ] now we choose @xmath1077 .
since clearly there exists a constant @xmath31 such that @xmath1078 for every @xmath1002 , we have @xmath1079\le \frac { c'}l f(\sigma),\end{aligned}\ ] ] where we used lemma [ th : diamante ] . plugging this inequality into , we find @xmath1080 which implies @xmath1081 and , together with , @xmath1082 choosing @xmath1083 we obtain and therefore .
@xmath1084 here we prove the lower bound @xmath179 for @xmath159 with @xmath116 large enough .
the result is particularly interesting in @xmath9 , in view of the matching upper bound in theorem [ th : d2gap ] , but as we mentioned the proof works also in @xmath1 .
actually , we give the proof in the three - dimensional case , which is slightly more complicated .
we introduce the matrix @xmath1086 , unitarily equivalent to the matrix @xmath87 ( the generator ) , as @xmath1087 the spectrum of @xmath438 coincides with the spectrum of @xmath1088 , so that we have to prove that the smallest non - zero eigenvalue of @xmath1089 is lower bounded by @xmath11 .
note that @xmath438 is symmetric thanks to the reversibility condition @xmath1090 .
the transformation is the analogue of _ the inverse _ of the `` ground - state transformation '' which maps a schrdinger operator of the form @xmath1091 ( which acts on @xmath1092 ) into the operator @xmath1093 where @xmath1094 is the ground state eigenfunction ( we assume for definiteness that @xmath1095 , i.e. the ground state energy is zero ) and the unitary operator @xmath262 acts as @xmath1096 for @xmath1097 .
note that @xmath1088 is the generator of a diffusion with drift .
looking at the definition of @xmath438 and @xmath1088 , one immediately realizes that * @xmath1098 * if @xmath1099 for some @xmath72 and @xmath1100 , then @xmath1101 * if @xmath1099 for some @xmath72 and @xmath1102 ( hence @xmath1103 ) then @xmath1104 * if @xmath1105 and there exists no @xmath22 such that @xmath1099 , then @xmath1106 .
if we write @xmath1107 with @xmath1108 we see that all the matrix elements of @xmath1109 are smaller than @xmath1110 if @xmath13 satisfies with @xmath116 large enough .
since each row of @xmath1109 has at most @xmath1111 non - zero elements , one sees easily that the spectral radius of @xmath1109 is @xmath1112 .
we are therefore left with the task of proving that the smallest non - zero eigenvalue of @xmath1113 is larger than @xmath11 .
thanks to formulas - , we have that @xmath1114 and , for @xmath1105 , @xmath1115 if we decompose @xmath19 into a finite number @xmath516 of equivalence classes @xmath1116 , where two configurations belong to the same class iff they can be connected via a finite number of single spin - flips which do not change the energy @xmath1117 , then @xmath1118 whenever @xmath1119 with @xmath1120 . in other words , one can write @xmath1121 in a block matrix form @xmath1122 where @xmath1123 and in particular , if @xmath1124 denotes the spectrum of @xmath1121 , one has @xmath1125 .
it is clear that , if @xmath1126 denotes the equivalence class whose unique element is the all `` @xmath0 '' configuration , one has @xmath1127 ( cf . and observe that any spin flip increases the energy if @xmath1128 ) .
we need to prove that , for every @xmath1129 , @xmath1130 for some positive @xmath31 .
let us fix @xmath1131 , let @xmath1132 be the smallest eigenvalue of @xmath1133 and @xmath1134 an associated eigenfunction .
we want to show that @xmath1135 with @xmath31 independent of @xmath886 .
the key point is the following : [ th : killed ] for every @xmath1136 and @xmath790 , one has @xmath1137 where @xmath1138 denotes the law of the ising evolution @xmath1139 in @xmath16 at @xmath37 with @xmath0 boundary condition , started from the configuration @xmath92 , and @xmath1140 is the random time @xmath1141 just check that the time derivatives of left- and right - hand side are the same .
@xmath1142 note that @xmath1143 is just the generator of the `` killed '' markov process which coincides with the ising evolution except that it is killed when it exits the set @xmath1144 , cf . .
since the matrix @xmath1145 has strictly positive entries for @xmath790 ( this follows from and the definition of the equivalence classes @xmath1146 ) , the perron - frobenius theorem implies that the eigenvalue @xmath1147 is non - degenerate and the eigenfunction @xmath1148 can be chosen strictly positive on @xmath1146 . normalizing @xmath1148 so that @xmath1149 , we have @xmath1150 the desired estimate then follows from lemma [ th : killed ] by letting @xmath105 , once we prove the following result : [ th : blocchi ] there exists @xmath1151 and @xmath7 independent of @xmath2 such that , for every @xmath1131 , @xmath1136 and @xmath790 one has @xmath1152 let @xmath515 be the smallest parallelepiped which contains the set @xmath1153 , and let @xmath672 be the rectangular layer of points of @xmath515 with maximal vertical coordinate . plainly , @xmath1154 where @xmath1155 this is because when the last `` @xmath3 '' spin in @xmath672 flips to `` @xmath0 '' , the energy decreases by at least @xmath1015 .
then , @xmath1156 where @xmath1157 is the configuration where spins take value `` @xmath3 '' in @xmath515 and `` @xmath0 '' in @xmath1158 , and we used monotonicity of the dynamics ( the event @xmath1159 is decreasing , and @xmath1160 ) . again by monotonicity , we can assume that @xmath515 is the entire domain @xmath16 , and that spins in @xmath1161 are frozen at the value `` @xmath3 '' during the entire evolution : both these operations make @xmath1162 stochastically larger .
but , in this case , @xmath672 is just a @xmath1163 square , and the evolution in @xmath672 coincides with the evolution of the two - dimensional ising model with @xmath0 boundary conditions and @xmath37 ( the `` @xmath0 '' boundary conditions on the upper face of @xmath672 and the `` @xmath3 '' boundary conditions on the lower face compensate exactly ) .
we have then @xmath1164 where @xmath1165 is the law of the evolution of the two - dimensional ising model in the @xmath1163 square with `` @xmath0 '' boundary conditions started from the `` @xmath3 '' configuration , and @xmath4 was defined in as the first time when all spins in the square are `` @xmath0 '' .
finally , one has @xmath1166^{\lfloor tc / l^2\rfloor } \le a(l)e^{-c\gamma \frac tl}\end{aligned}\ ] ] where in the first inequality we used monotonicity ( if @xmath4 has not been reached at time @xmath1167 , we restart the dynamics from the all `` @xmath3 '' configuration ) and in the second one the result of theorem [ th : d2 ] .
this ends the proof of theorem [ th : blocchi ] . @xmath1168
the asymptotic behavior @xmath1169 can be understood by comparison with the symmetric simple exclusion process ( ssep ) , as already noticed in @xcite .
in fact , by monotonicity , the tail of the law of @xmath4 is bounded from above by the tail of the hitting time @xmath1170 defined as follows .
consider the ssep on @xmath1171 $ ] starting with all the negative sites occupied and all the non - negative ones empty and define @xmath1170 to be the first time at which the site `` @xmath2 '' is occupied . in turn , using beautiful results by liggett and by arratia @xcite , the tail of @xmath1170 is controlled from above by the tail of the same hitting time but for the process in which the @xmath2 particles evolve as independent random walks without the exclusion constraint .
finally the latter has a tail @xmath1172 simply because the tail for a single random walk is @xmath1173 and there are @xmath2 of them .
we will prove here the upper bound @xmath1174 for the dynamics in the square box @xmath1175 .
we use the variational characterization of the spectral gap and we note that , if @xmath1176 is some subset of the configuration space @xmath19 , one has @xmath1177 where the infimum is taken over functions which vanish on @xmath1178 : just observe that @xmath1179 the bound @xmath1174 is therefore proven if one exhibits a set @xmath1176 such that @xmath1180 and a function @xmath102 which vanishes on @xmath1178 and such that @xmath1181 for @xmath68 , we let @xmath1014 and @xmath967 be as in definition [ def : gset ] and we let @xmath1182 be the square @xmath1183 ^ 2 $ ] .
we consider the eight points of @xmath251 which are at distance @xmath882 from a corner of @xmath249 and we call them @xmath1184 , with the convention that we order them clockwise , starting from the right - most one on the north side of @xmath249 .
let us assume for definiteness that @xmath2 is even .
with @xmath1185 and a configuration @xmath1186 : only the boundary @xmath1047 is drawn ( thick line ) .
the four portions @xmath1187 of @xmath1188 can not intersect and , if suitably rotated , are just lattice paths . according to formula , one has in this example @xmath1189 and @xmath1190.,scaledwidth=60.0% ] [ def : y ] we let @xmath1176 be the subset of configurations in @xmath19 such that @xmath967 is a simple curve of length @xmath1191 and such that @xmath1192 .
note that @xmath1191 is just the length of @xmath251 . also , if @xmath1193 then the following properties are immediately verified , cf . figure [ fig : paths ] : 1 .
the portion of @xmath1047 which connects @xmath1194 to @xmath1195 or @xmath1196 to @xmath1197 is a straight segment of length @xmath930 included in @xmath251 ; 2 .
the portion of @xmath1047 which connects @xmath1198 to @xmath1199 ( call it @xmath1200 ) is a `` lattice path '' of length @xmath2 ( in the sense of section [ app : wilson ] ) .
more precisely , if @xmath1200 is rotated counter - clockwise by an angle @xmath1201 , then expanded by a factor @xmath1004 and finally is suitably translated , then it becomes the graph of a function @xmath1202\mapsto \phi^{(i)}_x:=\phi^{(i)}_x(\sigma)\in[-l/2,l/2]$ ] such that @xmath1203\\ \label{eq : uu } \phi^{(i)}_\cdot \mbox{\;is linear in each interval\ ; } ( x , x+1 ) , x\in { { \mathbb z}}\cap [ -l/2 , l/2 - 1].\end{gathered}\ ] ] the test function @xmath102 in is then @xmath1204 where @xmath1205 with the convention that @xmath1206 .
note that @xmath102 vanishes on @xmath1207 and that @xmath1208 as required above .
let @xmath445 be the uniform measure over the set of lattice paths @xmath310 satisfying - , and let @xmath1209 be the generator @xmath1210.\end{aligned}\ ] ] [ rem : otimes ] conditionally on @xmath1193 , the four lattice paths @xmath1211 are independent and , since all their allowed configurations have the same length , @xmath1212 is just the product uniform measure @xmath1213 , where each copy of @xmath445 acts on a path @xmath1214 .
indeed , the ising hamiltonian equals @xmath1215 which does not depend on @xmath29 if @xmath1193 , cf . definition [ def : y ] .
[ th : lemmapiro ] if @xmath102 is defined as in , one has @xmath1216
let us consider first the denominator of the left - hand side : @xmath1217 ^ 4\end{aligned}\ ] ] where we used the fact that @xmath1218 , see remark [ rem : otimes ] .
the numerator requires more work . from the definition of the generator one has @xmath1219 where @xmath1220 is defined in .
the first observation is that , if @xmath159 with @xmath116 large enough , then @xmath1221 whenever @xmath1193 and @xmath212 is such that the configuration @xmath1222 satisfies @xmath1223 ( in particular , @xmath1222 does not belong to @xmath1176 ) .
the reason is that , since @xmath1224 , one has @xmath1225 where @xmath1226 is the probability , under @xmath1227 , that the spin at @xmath212 equals @xmath1228 , and this probability is smaller than @xmath1229 by the assumption that @xmath1223 ( cf . ) .
call @xmath1230 the set of @xmath1231 such that @xmath1232 .
one has then @xmath1233 since , if @xmath1193 , there are no spin flips which decrease @xmath1234 ( i.e. the energy ) . finally , recalling remark [ rem : otimes ] it is an easy task to check the identity @xmath1235 ^ 3 \rho(g(-\hat { \mathcal l})g),\end{aligned}\ ] ] the reason being that the flips which leave @xmath1234 unchanged just correspond to the local updates in the generator @xmath1209 of .
the factor @xmath1236 is due to the symmetry among the lattice paths @xmath1214 .
@xmath1237 the proof of is concluded once we prove [ th : ensembles ] there exists @xmath7 such that @xmath1238 using the form of the generator and the fact that @xmath445 is the uniform measure over lattice paths , one sees that @xmath1239 ^ 2}{\cos^2(\pi x / l)}\rho\left(g^2{\bf 1}_{\{\nabla_{x-1}=+1,\nabla_x=-1\}}\right)\\ \nonumber & & + \frac12\sum_{x =- l/2 + 1}^{-1}\frac{\left [ \cos(\pi ( x-1)/l)-\cos(\pi
x / l ) \right]^2}{\cos^2(\pi ( x-1)/l)}\rho\left(g^2{\bf 1}_{\{\nabla_{x-1}=-1,\nabla_x=+1\}}\right)\end{aligned}\ ] ] where we recall that @xmath1240 and we used the symmetry @xmath1241 to restrict the sum to @xmath1242 . clearly , one has @xmath1243
2\le c / l . \end{aligned}\ ] ] moreover , by an `` equivalence of ensembles '' argument , one has @xmath1244 and a similar estimate for @xmath1245 where @xmath917 is finite uniformly in @xmath1246 .
the statement of the theorem then immediately follows from eqs . , and .
the proof of is a direct adaptation of the proof of ( * ? ? ?
3.8 ) . loosely speaking ,
the idea is to observe that the law of @xmath1247 under @xmath1248 is a product law of independent ( but not identically distributed ) random variables conditioned to the event @xmath1249 .
next , one shows that , when computing the average of a local function like @xmath1250 , the conditioning can be eliminated , at the price of losing a multiplicative constant . under the unconditioned product law ,
the expectation of @xmath1250 is easily computed and turns out to be proportional to the r.h.s . of .
we wish to thank rick kenyon for very helpful e - mail exchanges about dimer coverings and height fluctuations and david wilson for a precious suggestion on the proof of the mixing time lower bound in three dimensions .
this work has been carried out while f. simenhaus and f. l. toninelli were visiting the department of mathematics of the university of roma tre under the erc advanced research grant `` ptrelss '' .
they both gratefully acknowledge the kind hospitality and support . | we consider the glauber dynamics for the ising model with `` @xmath0 '' boundary conditions , at zero temperature or at temperature which goes to zero with the system size ( hence the quotation marks in the title ) . in dimension @xmath1 we prove that an initial domain of linear size @xmath2 of `` @xmath3 '' spins disappears within a time @xmath4 which is at most @xmath5 and at least @xmath6 , for some @xmath7 .
the proof of the upper bound proceeds via comparison with an auxiliary dynamics which mimics the motion by mean curvature that is expected to describe , on large time - scales , the evolution of the interface between `` @xmath0 '' and `` @xmath3 '' domains .
the analysis of the auxiliary dynamics requires recent results on the fluctuations of the height function associated to dimer coverings of the infinite honeycomb lattice .
our result , apart from the spurious logarithmic factor , is the first rigorous confirmation of the expected behavior @xmath8 , conjectured on heuristic grounds @xcite . in dimension @xmath9
, @xmath4 can be shown to be of order @xmath10 without logarithmic corrections : the upper bound was proven in @xcite and here we provide the lower bound . for @xmath9 , we also prove that the spectral gap of the generator behaves like @xmath11 for @xmath2 large , as conjectured in @xcite .
+ + 2000 _ mathematics subject classification : 60k35 , 82c20 _ + _ keywords : ising model , mixing time , dimer coverings , glauber dynamics . _ |
active galactic nuclei ( agns ) are among the most energetic objects in the universe , with energy output spanning the entire electromagnetic spectrum .
it is currently believed that the majority of soft x - ray emission by agns arises from superheating of the accretion disk surrounding the central supermassive black hole . however , the x - ray emission actually observed from agns depends strongly on the degree to which the `` central engine '' is obscured by surrounding material ( e.g. , the dusty torus model )
. the observed spectrum ( i.e. , hardness ) of the source also depends on the amount of obscuration present .
this paper seeks to study the soft x - ray properties of cosmologically local ( @xmath11 ) agns . towards this end
, we have constructed an optically selected sample of such agns detected as part of the rosat all - sky survey ( rass ) . in order to better understand the x - ray emission properties of local agns ,
we have compiled a relatively unbiased sample of these objects for study . the kpno international spectroscopic survey ( kiss ; salzer et al .
2000 , 2001 ) provides a unique sample of galaxies ideally suited for statistical analyses .
the kiss sample includes all emission - line galaxies ( elgs ) with detectable h@xmath12 emission ( ew @xmath13 ) .
since it is a line - selected sample with an imposed maximum wavelength , there is an implicit redshift limit associated with the survey .
this limit , along with the great sensitivity of the survey , means that kiss is effectively volume - limited for the more luminous portion of the elg sample , which includes the majority of the agns . because of the unique way in which kiss galaxies
are selected , our rosat - kiss comparison differs in a fundamental way from previous x - ray / optical correlation studies .
most of these are x - ray selected ( e.g. , ueda et al . 1999 , griffiths et al . 1996 ) and focus on the optical properties of objects detected in the x - ray . here
, we are studying the x - ray properties of an optically selected sample .
this approach yields a sample of galaxies significantly different from those derived by previous authors ( see section [ properties ] ) .
studies by @xcite and @xcite also use optical catalogs as the primary basis for selection .
their galaxy samples are drawn from the palomar survey of nearby galaxies , a magnitude - limited survey which has full coverage of the northern sky .
the sample constructed from this survey is volume - limited and complete to a comoving distance of @xmath14 18 mpc .
although the sky coverage of kiss is only 128 @xmath15 in total , the sample volume is quite deep : the comoving distance limit of the survey is @xmath16 mpc .
the current study complements those of halderson et al .
and ho et al . by exploring the x - ray properties of a much more distant ( but still cosmologically `` local '' ) sample . on average ,
the rosat - kiss sample is @xmath14 100 times more luminous than the halderson et al .
sample in the soft x - ray band .
emission - line galaxies similar to those discovered by kiss are important components of the cosmic x - ray background ( xrb ) .
the xrb was discovered 40 years ago @xcite in the 26 kev range , and has since been demonstrated to extend to both higher and lower energies ( see the review by fabian & barcons 1992 ) . in deep images obtained recently with the _ chandra x - ray observatory _
, @xmath17 8090% of the background in the soft ( 0.52 kev ) and hard ( 210 kev ) x - ray bands has been resolved into discrete sources ( e.g. , rosati et al .
models ( comastri et al .
1995 ; gilli et al.2001 ) and observations of both nearby ( moran et al .
2001 ) and distant ( tozzi et al .
2001 ) x - ray sources have indicated that the majority of the hard xrb is produced by type 2 agns whose soft x - ray fluxes are absorbed by dense circumnuclear material . in contrast , at least 50% of the xrb at lower energies is produced by high - luminosity , unobscured ( type 1 ) agns and quasars ( hasinger et al . 1998 ; schmidt et al .
1998 ) . a significant fraction of the remaining soft xrb appears to be associated with optically bright galaxies of relatively low x - ray luminosity ( hornschemeier et al . 2001 ; tozzi et al .
it is unclear whether these objects are low - luminosity agns , non - active galaxies , or a combination of the two . a major aim of this study , therefore , is to constrain the soft xrb contribution of moderately luminous agns .
we describe the kiss and rosat catalogs and the results of the x - ray - optical matching in section 2 .
the properties of the galaxies in the correlation sample are discussed in detail in section 3 .
we present xrb calculations in section 4 ; we take into account contributions from both rosat - kiss objects and those whose 0.1 - 2.4 kev emission lies below the rosat detection threshold .
we close with a discussion of the implications of our analysis in the context of current xrb research .
kiss is a line - selected , objective - prism survey for extragalactic emission - line objects @xcite .
kiss is the first purely digital spectroscopic survey for emission - line galaxies ( elgs ) .
it therefore represents a significant advance beyond slitless spectroscopic surveys carried out using photographic plates .
examples of such photographic surveys include the markarian @xcite , michigan ( um ; @xcite ) , case @xcite and ucm @xcite surveys .
its digital nature allows kiss to reach much fainter flux limits than previous studies , and therefore to reach a substantially greater galaxian surface density ( 18.1 @xmath18 over the @xmath5 strip ; see below ) .
kiss is being carried out over several `` strips '' , each @xmath19 wide in declination @xcite .
the first survey area ( kr1 ; hereafter the @xmath5 strip ) covers the region ra = 12h 15 m - 17h 0 m , @xmath20 = @xmath21 - @xmath5 , coinciding with part of the cfa century redshift survey @xcite .
the second ( kr2 ; hereafter the @xmath6 strip ) runs through the bootes void : ra = 11h 55 m - 16h 15 m , @xmath20 = @xmath22 - @xmath23 .
the total sky coverage of these two survey regions is 128.1 @xmath15 . in the @xmath5 strip , 1128 elgs
have been cataloged @xcite , while a total of 1029 have been identified in the @xmath6 strip @xcite . to date , high - resolution follow - up spectra have been obtained for 64% of the objects in the @xmath5 strip and 14% of those in the @xmath6 strip , allowing precise determination of redshift and elg activity type ( i.e. , starburst vs. agn ) .
survey data were obtained in two distinct spectral regions . the blue survey ( salzer et al .
2002 ) selects elgs by the presence of the [ o iii ] @xmath245007 emission line ; the red survey ( see salzer et al .
2001 ) selects _ via _ the h@xmath12 line .
this paper focuses on the red spectral data : only the h@xmath12-selected portion of the kiss survey data is considered .
the spectral range of the red survey data extends from 6400 to 7200 , beginning slightly below the rest wavelength of the h@xmath12 line and cutting off before the oh molecular band at about 7240 .
this implies that the h@xmath12 line is detectable to a redshift of @xmath25 . in a minority of cases
a high - redshift emission - line object will be detected by kiss when another strong spectral line ( e.g. , [ o iii ] , h@xmath26 ) redshifts into the kiss wavelength range .
several qsos and seyfert galaxies have been detected by kiss in this manner .
because of its well - defined redshift cutoff , kiss is essentially volume - limited for the more luminous portion of the sample .
some incompleteness arises for galaxies with faint h@xmath12 line strengths .
however , the majority of the kiss agns are included in the volume - limited portion of the sample . for a detailed discussion of kiss completeness limits ,
see @xcite .
due to its depth and emission line selection method , kiss is ideal for performing relatively unbiased statistical studies of the local population of star - forming galaxies and agn .
the rass resulted in a sample of soft ( 0.1 - 2.4 kev ) x - ray sources which is a factor of 20 deeper than previous x - ray surveys such as the einstein medium sensitivity survey @xcite , and which has nearly full sky coverage . at a brightness limit of 0.1 cts
/ s , the survey covers 92% of the sky @xcite .
the large effective area and short focal length of rosat , combined with the low background noise of the pspc , resulted in an instrument that was designed to study x - ray point sources , but which was nearly ideal for studying diffuse x - ray emission as well . as a result ,
the survey yielded the most accurate maps to date of the diffuse xrb in the 0.1 - 2.4 kev energy band @xcite .
these maps have an angular resolution of approximately @xmath27 .
the resolution of these maps was later improved to 12 @xcite .
deep surveys later conducted by rosat @xcite revealed that 70 to 80% of this diffuse background could in fact be resolved into discrete sources to flux levels on the order of @xmath28 @xmath29 @xmath30 @xmath31 .
the rass was performed using the position sensitive proportional counter ( pspc ) , an instrument with a positional accuracy on the order of 10 arcsec ( correlation with the tycho catalog of bright stars shows that 68% of bright source catalog objects are found within 13 arcsec of the optical position ; voges et al . 1999 ) .
catalogs derived from rass provide detailed information about each source , including position , positional error , source count rate and error , and exposure time .
in constructing the rosat - kiss database , we have combined all relevant x - ray source information with the optical data contained in the kiss tables @xcite .
the x - ray data used in this paper is drawn from the bright and faint source catalogs , which are derived directly from processing of the all - sky survey data @xcite .
the bright source catalog ( bsc ) consists of the brightest objects in the rass - ii database ( resulting from the second rass processing ; see voges et al .
the limiting pspc count rate is @xmath32 cts / s with at least 15 source counts required for inclusion , or a detection likelihood of at least 15 .
detection likelihood is defined as @xmath33 , where p is the probability that the observed photon distribution is due to random background fluctuations .
these two criteria result in a source catalog of 18,811 objects .
the second and larger catalog constructed from all - sky survey data is the faint source catalog ( fsc ) .
the fsc is an extension of the bsc to fainter limiting count rates ; it represents the full measure of the non - spurious sources detected in the rass - ii processing , containing 145,060 objects .
the criterion for inclusion in the fsc is at least 7 source counts , with a detection likelihood of 7 or greater .
the median count rate for fsc sources in the two kiss areas is approximately @xmath34 cts / s .
( see section [ predictions ] for discussion of the fsc flux threshold . )
previous rass correlations have been performed with catalogs of sources at various wavelengths .
these include correlations with the rosat catalog of pointed observations ( wgacat ; white et al . 1994 ) , the simbad and ned public databases , the first radio survey @xcite , and the iras catalogs @xcite .
a positional - coincidence test was used to conduct the correlation between the kiss database and the bsc / fsc , based on correspondences between the coordinates of objects in x - ray and optical wavelengths .
a uniform search radius of 60 arcsec was adopted : all rosat sources within 60 arcsec of a kiss elg were considered to be possible matches . the median formal uncertainty in rosat source position recorded in the bsc / fsc for rosat - detected kiss objects is @xmath14 10 arcsec , with errors as large as 25 arcsec present in the final dataset .
there were a total of 22 matches in the first - pass correlation table , which includes all matches , legitimate and spurious , for rosat bsc and fsc sources appearing within 60 arcsec of a kiss object .
the median difference between x - ray and optical source positions was 10.9 arcseconds , with a standard deviation of 12.0 ; any source with a positional offset greater than 30 arcseconds was therefore flagged as a possible spurious match .
each source pair was verified visually using the skyview database , to rule out chance correspondences .
x - ray contour images drawn from the archive and overlaid on the online digitized sky survey ( dss ) images were used to confirm positional matchesr extraterrestrische physik ( mpe ) at garching , germany . ] . for a number of sources
the kiss object is too faint to be visible on the dss images . in these cases
the location of the x - ray contours relative to other objects in the dss field was compared with the finder charts available for all kiss elgs ( which are substantially deeper than the dss images ) in order to verify the match .
three kiss emission - line galaxies with positional offsets of greater than 30 arcsec were discarded after visual examination ( kissr 624 , 662 , and 2128 , with 57.2 , 33.7 and 48.8 arcsec positional offsets , respectively ) .
another matched galaxy , kissr 229 , was found not to be an elg after acquisition of a follow - up spectrum .
furthermore , the positional offset between the rosat source and kissr 229 is 24.7 arcsec .
thus , kissr 229 is not likely to be the true optical counterpart for this x - ray source , since the high density of non - elgs in these fields greatly increases the odds of chance correspondence with an x - ray source position .
the remaining 18 objects were judged to be likely matches after visual examination .
the median x - ray / optical positional offset for these galaxies is 9.6 arcsec .
this is roughly consistent with the results of the correlation by brinkman et al .
( 2000 ) of the rosat and first ( 20 cm ) catalogs .
this study finds that only 2% of all rosat - first matches are chance coincidences , to an angular separation of 30 arcsec .
the 18 objects described above comprise the rosat - kiss sample .
their distribution between the two kiss areas and the bsc / fsc are summarized in table [ table1 ] , while their optical and x - ray characteristics can be found in table [ table2 ] .
comparisons were also made between the kiss catalogs and the rosat public archive of pointed pspc observations ( wgacat ) .
the correlation was conducted remotely via the heasarc server , again using a search radius of 60 arcsec .
only two rass - detected sources were found to have corresponding pointed observations : kissr 281 , a seyfert 1 galaxy , and the nearby liner kissr 29 ( ngc 4278 ; for details , see section [ properties ] ) .
observations of the latter object can also be found in the public archive of rosat high resolution imager ( hri ) images ; further search of this latter catalog revealed that no other kiss galaxies had been detected .
for the sake of uniformity , we use only the rass pspc fluxes for these two sources in our analysis .
-0.5 in = 7.0 in -0.3 in -0.5 in = 7.0 in -0.3 in -0.5 in = 7.0 in -0.3 in 0.2 in = 4.4 in 1.3 in
the rosat - kiss sample consists of 18 x - ray detected sources .
the basic properties of these objects are listed in table [ table2 ] .
column 1 gives the kissr catalog number of each source , while column 2 lists the corresponding designation in the appropriate rass catalog .
columns 3 and 4 list the ra and dec coordinates of each source ( equinox j2000 ) , taken from the kiss catalogs @xcite .
the symbol @xmath35 in column 5 of the table represents the offset in arcseconds between the optical and x - ray positions of the object . in column 6 ,
the elg type is given for each object .
the absolute and apparent b magnitudes of all rosat - kiss sources are listed in columns 7 and 8 .
the b - v color of all objects is recorded in column 9 , while their redshifts can be found in column 10 .
all photometric and spectroscopic data come from the kiss database .
the pspc count rate for each source is given in column 11 , while the 0.1 - 2.4 kev luminosities derived from this count rate ( using appropriate spectral parameters ; see section 4.1 ) make up column 12 . follow - up optical spectra have been obtained for approximately 900 galaxies in the overall kiss sample , during the period from 1998 - 2001 .
results of spectral observations are given in @xcite , @xcite , and @xcite .
the spectroscopic properties of the kiss elg sample as a whole can be found in @xcite .
follow - up optical spectra are available for all 18 rosat - kiss objects , making activity type classifications ( agn / starburst ) and accurate redshift determination possible in all cases .
the spectra for all objects in the rosat - kiss sample are shown in figure [ figure1 ] .
knowledge of the spectroscopic properties of the sample is extremely useful for establishing the nature of each source .
accurate redshift data is also essential in xrb calculations requiring knowledge of the emissivity within a certain volume ( section [ xrb ] ) , where it becomes necessary to exclude luminosity contributions from sources outside the limiting h@xmath12 redshift of kiss .
the spectral characteristics of the x - ray selected sample vary widely , from stellar absorption lines to broad - line qso emission .
thirteen of the 18 objects in the rosat - kiss sample have been spectroscopically identified as seyfert 1 galaxies .
this high percentage of seyfert 1s is most likely due to the fact that the rosat pspc detection threshold corresponds to a relatively high flux level , meaning that only the brightest soft x - ray sources will be reliably detected .
the low degree of obscuration associated with broad - line objects implies that they will be more readily detected in the soft x - ray band than will narrow - line agns .
four of the remaining five objects are other types of agn , including one seyfert 2 galaxy , one qso , and two liners .
the remaining source , kissr 817 , is identified as a stellar x - ray source ( see section [ indiv ] ) .
we note in passing that although kiss contains a sample of @xmath14 2000 starbursting galaxies , none of these objects were detected by rosat .
figures [ figure2 ] and [ figure3 ] show the distribution of apparent and absolute magnitudes for both the general kiss sample and the rosat - kiss sample .
the median in each case is brighter for the rosat - detected objects than for the general kiss objects . for the kiss sample as a whole , the median absolute magnitude @xmath36 .
this is more than a full magnitude less luminous than the median for the rosat - kiss sample ( @xmath37 ) .
the same is true for apparent magnitude , where the median @xmath38 for the general kiss population , but is @xmath39 for the rosat - kiss objects .
the general kiss elg sample is heavily dominated by starbursting galaxies .
for example , @xmath40 of the 663 kiss elgs with follow - up spectra from the @xmath5 strip are starbursting galaxies .
only 5.3% of these galaxies are seyferts , and 5.7% are liners . of the seyfert galaxies , 34.3% exhibit broad lines ( seyfert 1s ) .
there are only 19 seyfert 1 galaxies identified to date in the whole of the sample . a large fraction ( 68% ) of the seyfert 1s detected by kiss
belong to the rosat - detected sample .
these galaxies are among the most optically luminous objects in the kiss database ( median @xmath41 ) .
an interesting aspect of the rosat - kiss seyfert 1 subsample is the high proportion of `` narrow - line '' seyfert 1 ( nls1 ) galaxies present .
inspection of figure 1 reveals that many of the seyfert 1 s exhibit small fwhm values for both h@xmath26 and h@xmath12 , as well as substantial fe ii emission on either side of the h@xmath26 and [ o iii ] lines .
examples include kissr 281 , 350 , 489 , 531 , 1511 , 1755 , and 1903 .
these galaxies all have h@xmath26 fwhms of less than 1500 km / s and [ o iii]/h@xmath26 flux ratios of less than @xmath42 , consistent with the emission - line criteria for nls1 s described by osterbrock & pogge ( 1985 ) and goodrich ( 1989 ) .
nls1 s have generally steeper soft x - ray spectra than classical seyfert 1 s ( boller , brandt , & fink 1996 ) ; as a result , previous soft x - ray selected samples of agns have included high fractions of these objects ( e.g. , stephens 1989 ; bade et al .
1995 ; grupe et al .
1999 ) . 0.2 in = 4.4 in 1.3 in remarkably ,
not only does a large fraction of the rosat - detected seyfert 1 sample of kiss galaxies exhibit narrow lines ( 7 out of 13 , or 54% ) , but _ a large fraction of all kiss seyfert 1 galaxies are nls1s_. in total there are 16 seyfert 1 galaxies in the kiss catalog within the h@xmath12-selected redshift region ( @xmath43 ) . of these ,
at least 7 ( 44% ) would qualify as nls1 s .
while the fraction of sources among the general population of seyfert 1 galaxies possessing narrow lines is not well known , the numbers for the kiss sample ( which is selected in a relatively unbiased way , at least with regard to line widths ) suggest that it may be quite high .
further analysis of this issue will await more complete spectroscopic follow - up of the kiss elgs . in order to find the x - ray luminosities of the objects in the rosat - kiss sample , we first must convert rass count rates to fluxes in the 0.1 - 2.4 kev band . to do this ,
we apply an absorbed power - law model with @xmath44 galactic + @xmath45 @xmath31 to convert pspc source counts to x - ray flux .
these spectral parameters are consistent with the methods of @xcite , who determined @xmath44 galactic + @xmath45 @xmath31 through modeling of rosat pspc spectra for a low - luminosity , distance - limited agn sample . @xcite use @xmath46 , with @xmath47 equal to the galactic value , for flux calculations ( found in the electronic form of the rass correlation catalogs ) .
we make use of pimms in performing the conversion between rosat pspc source counts and the @xmath1 kev flux .
combining the calculated x - ray fluxes from the above spectral model with the distances calculated from measured redshifts allows us to calculate the absorption - corrected @xmath1 kev luminosity of each source .
this information is given in table [ table2 ] .
we use the _ _ colden _ _ software package , in which column densities are drawn from the nrao dataset @xcite , to calculate the galactic h i column density along the line of sight to each galaxy . for the rosat - kiss objects in the @xmath5 strip
, the galactic component of @xmath47 has a median value of @xmath48 @xmath31 ; @xmath6 strip objects have a median galactic component of @xmath49 @xmath31 .
0.2 in = 4.4 in 1.3 in knowledge of the total x - ray emissivity over the kiss volume is necessary for an understanding of the xrb contribution made by the low - luminosity agns in the kiss sample ( section 4 ) .
however , the majority of the kiss sample is not detected by rosat . since the optical luminosities of the rosat - kiss sample are not drastically different from those of the general sample ( section [ properties ] ) , we expect that the contribution to the xrb made by kiss objects below the rosat threshold will be non - negligible .
this implies that an estimate of the x - ray luminosity of the non - detected kiss galaxies is necessary for an accurate estimate of the xrb contribution from the full kiss agn sample . in an attempt to constrain the x - ray luminosities of the kiss seyfert galaxies not included in the rosat - kiss sample ,
we derive an @xmath50 relation using the 15 lower redshift extragalactic objects for which x - ray data are available .
the h@xmath12 luminosities of these galaxies are calculated using the objective - prism spectrum of each galaxy .
the objective - prism fluxes are used , rather than the values from the follow - up slit spectra , because the former are calibrated in a uniform , homogeneous way ( see kr1 and gronwall et al .
since the @xmath0 and [ n ii ] ( @xmath516548 ) lines are blended in the objective - prism spectra , the observed [ n ii]/@xmath0 ratio from the follow - up spectra are used to correct the objective - prism fluxes for the contribution due to [ n ii ] .
in addition , the fluxes of the seyfert 2 and liner galaxies are corrected for intrinsic absorption using reddening coefficients calculated from the balmer decrement .
this latter correction can be especially significant for the liners in our sample .
the resulting @xmath50 relation is shown in figure 4 . in the figure
seyfert 1 galaxies are plotted as filled squares , the lone seyfert 2 as a filled triangle , and the two liners as open triangles .
there appears to be a fairly well - defined linear relation between @xmath8 and @xmath52 for these galaxies , at least for the seyfert 1s .
the median @xmath7 ratio for the volume - limited sample is 20.7 , which is illustrated by the solid line in figure 4 .
although a significant portion of the kiss agns have a soft @xmath8 too faint to allow detection by rosat , this does not eliminate the possibility that these objects may be responsible for a non - trivial xrb contribution . 0.2 in = 4.4 in 1.3 in
as mentioned above , many kiss agns have not been detected by rass . in order to derive an accurate assessment of the contribution of kiss - type agns to the soft xrb , the x - ray luminosities of the undetected objects must be estimated .
to do this , we employ the l@xmath53/l@xmath54 ratio and the measured h@xmath12 luminosities of the galaxies .
our initial plan had been to utilize the value derived in the previous subsection for the rosat - kiss sample .
however , examination of figure 4 shows that our median value for the @xmath7 ratio is dominated by the seyfert 1 galaxies in the sample , while the seyfert 2s and liners are systematically lower .
since the vast majority of non - rosat - detected objects are seyfert 2s and liners , we wanted to use an @xmath7 ratio that was representative of these classes of agn .
however , we did not feel comfortable using the small sample of narrow - lined rosat - kiss galaxies ( three ) to determine a median @xmath7 ratio to be used in our computations .
therefore , we decided to adopt the @xmath7 ratio determined by @xcite for their local sample of mainly narrow - lined agns .
these authors found a median @xmath7 of 7 .
it is no great surprise that the median @xmath7 we calculate for the rosat - kiss sample is larger than that found by @xcite , since our sample is dominated by seyfert 1 galaxies . because of the greater obscuration expected for type 2 seyferts in the dusty torus model , unabsorbed emitters such as seyfert 1s
are expected to have a significantly higher @xmath7 in the soft x - ray band .
this is precisely what we observe in our sample .
one can not simply use the value of the @xmath7 ratio to estimate @xmath8 .
examination of the distribution of fsc sources with flux reveals a loss of completeness beginning at approximately @xmath55 cts / s .
we therefore adopt this value as an upper limit on the pspc count rate for our non - detected objects .
computing @xmath8 using the average @xmath7 ratio described above will , in some cases , lead to an estimate of the x - ray flux for an object which lies above the fsc flux limit .
we therefore adopt the upper limit @xmath1 kev count rate of 0.015 cts / s for all such objects , rather than the value inferred from the @xmath7 ratio .
this method yields , in turn , an upper limit for the soft xrb contribution of these non - rosat - detected kiss galaxies .
figure [ figure5 ] shows the @xmath56 relation with the non - rosat kiss galaxies included .
the non - detected seyfert galaxies are plotted as filled circles , while non - detected liners are open circles .
objects whose @xmath8 estimate is based on the fsc flux limit , as descibed above , are indicated by an upper - limit arrow .
many of the non - detected seyfert galaxies fall into this group .
all other non - detections have had their x - ray luminosities predicted using the relation @xmath57 .
these objects , which include many of the liners , fall along a straight line in the plot .
there are a total of 65 seyfert galaxies in the kiss database .
fourteen of these galaxies are detected by rosat , leaving 51 seyferts with no x - ray flux data .
after elimination of those galaxies outside the kiss completeness limit ( @xmath58 ; see section 4.1 ) , 36 non - rosat - detected seyfert galaxies remain .
kiss also detects a large population of liners - there are 55 liners in the combined @xmath5 and @xmath6 kiss database .
most of these objects have not been detected by rosat , and nearly all of them lie within the @xmath58 completeness limit ; we predict @xmath8 using the above method for 52 liners in total . since the rosat - kiss sample consists of just 18 objects , it is possible for us to examine each one in some detail , while still making generalizations about the sample as a whole .
this section details the properties of selected rosat - kiss objects .
_ kissr 29 : _ there are only two liners in the rosat - kiss sample ( the other example being kissr 2059 ) .
kissr 29 ( ngc 4278 ) is remarkable in that it is one of the optically brightest sources in the kiss survey ( @xmath59 ) ; this galaxy is extremely nearby ( only 8 mpc distant ) .
kissr 29 is the only rosat - kiss object for which archival hri images are available .
however , in the interests of uniformity we have not made use of the pointed data in our calculations .
@xcite use pointed pspc observations of kissr 29 to derive an estimate of the total soft x - ray flux ( 0.1 - 2.4 kev ) of @xmath60 ergs @xmath30 @xmath31 , assuming a photon index of @xmath61 and column density @xmath62 galactic + @xmath63 @xmath31 . using the same spectral parameters , we find that the 0.1 - 2.4 kev flux for this object is @xmath64 ergs @xmath30 @xmath31 using the rass detection .
_ kissr 817 : _ this is the only stellar x - ray source in our sample .
kissr 817 is believed to be a cataclysmic variable star serendipitously detected by the all - sky survey ; it was detected as part of our sample due to its relatively high soft x - ray flux .
the follow - up optical spectrum detected no @xmath65 emission .
its spectrum resembles that of a hot white dwarf , and it is quite blue in color ( @xmath66 ) .
this is consistent with the behavior of a cataclysmic variable ; the star may have been in a high state when the objective - prism survey data were gathered .
this would explain the presence of the emission line in the survey data , of which no evidence was seen in follow - up observations .
_ kissr 864 : _ this object is one of the five qsos identified to date in the kiss catalog . as with all high - redshift kiss objects , it was detected via the presence of a strong emission line other than h@xmath12 which had redshifted into the relevant wavelength range . in this case , the emission line was @xmath67 .
kissr 864 is the most luminous object in the rosat - kiss sample , both in the optical ( @xmath68 ) and in the x - ray ( @xmath69 ) . _
kissr 1494 : _ one kiss seyfert 2 galaxy , kissr 1494 , was detected by the all - sky survey .
this galaxy is one of the more luminous sources in our rosat sample , with an absolute magnitude of -20.89 .
the median absolute magnitude for the overall sample of 46 kiss seyfert 2 galaxies is @xmath70 .
_ kissr 1802 _ : this object is the only high - redshift seyfert 1 galaxy detected by rosat - kiss .
the h@xmath26 and [ o iii ] emission lines are redshifted into the kiss bandpass , revealing a redshift of @xmath71 . as one might expect , this object is a very powerful x - ray source , with @xmath72 , making it the second most luminous source in the rosat - kiss sample .
_ seyfert 1 objects : _ for our purposes , all galaxies exhibiting broad balmer emission lines and with @xmath73 are termed seyfert 1 galaxies .
this includes a few `` intermediate '' seyferts , such as the seyfert 1.5 galaxies kissr 1503 and kissr 1802 .
the thirteen seyfert 1 galaxies in the rosat - kiss sample have a very high median luminosity : @xmath74 as compared to the general median @xmath75 for all kiss objects .
the median @xmath76 for the rosat - kiss sy 1 s is statistically indistinguishable from the value for the general kiss sy 1 sample .
as discussed above , the proportion of narrow - line seyfert galaxies ( h@xmath26 fwhm @xmath77 km / s , [ o iii]/h@xmath78 3 ) in the rosat - kiss sample seems to be significantly higher than that found in previous optically - selected agn samples .
determination of the relative contributions of various source populations to the x - ray background is of major importance . as summarized in section 1
, a modest fraction of the soft xrb may well be due to a population of moderate - luminosity agns . in this section ,
we perform comoving volume and emissivity calculations in order to estimate the xrb contribution from the types of agns detected locally by kiss .
we assume a cosmology of @xmath79 @xmath80 , @xmath81 , @xmath82 for all computations .
computation of the total volume covered by kiss ( necessary for volume emissivity determination ; see section [ xrb ] ) requires us to set a maximum redshift for the survey . the redshift cutoff for kiss objects does not occur instantaneously : we experience significant `` edge effects '' for objects near the cutoff redshift , due to loss of flux from the relevant emission line ( h@xmath12 ) .
studies @xcite indicate that this effect begins to become significant at approximately @xmath83 .
we therefore do not consider any kiss objects with spectroscopic redshifts larger than @xmath58 .
this results in elimination of one source ( kissr 1903 ; z=0.094 ) .
in addition , we do not consider the two high - redshift objects noted in section [ indiv ] ( kissr 864 and 1802 ) during computation of volume emissivities .
edge effects are present not only at the outskirts of a survey volume , but at the inner limit as well ; the distribution of extremely nearby sources will be heavily influenced by the local structure of the universe .
we have therefore excluded from our calculations the only rosat - kiss source close enough to experience these effects - the liner , kissr 29 ( ngc 4278 ) . after the elimination of all sources outside the relevant volume , we are left with a sample of 13 galaxies .
this sample , selected from a nearly volume - limited survey , should be representative of the local x - ray - emitting galaxy population as a whole .
accurate estimates of the emissivity produced by the objects in the kiss volumes are necessary to obtain a knowledge of the xrb contribution from these objects .
once the emissivity over the actual volume is known , it can be generalized to all redshifts using the appropriate evolution model ( see below ) in order to estimate the x - ray intensity over the sky area in question , which can then be compared with the known intensity of the xrb .
the areal coverage of the kiss strips is , in fact , fairly small : the @xmath5 strip covers an area of 62.16 @xmath84 , while the @xmath6 strip covers 65.86 @xmath84 .
the volume encompassed by the survey is nevertheless quite large relative to magnitude - limited surveys , because of the comparatively large redshift limit .
the comoving distance limit corresponding to our adopted @xmath85 is @xmath14 343 mpc .
this leads to volume emissivities in the @xmath1 kev band of @xmath86 ergs @xmath30 @xmath87 for the @xmath5 strip and @xmath88 ergs @xmath30 @xmath87 for the @xmath6 strip .
the quoted errors are simply poissonian , based on the numbers of objects detected in the two survey strips .
we have opted to employ poisson statistics in computing our uncertainties since small number statistics will dominate over the uncertainties in deriving the x - ray luminosities for the individual objects .
our approach leads to more conservative estimates of the uncertainties .
it is worthwhile to note that the volume emissivities from the two kiss strips differ by a factor of over four .
this is largely a result of the small number of objects in each sample .
in fact , much of the difference between the xrb contributions from the rosat - kiss objects in the two strips is due to a single , high - luminosity source in the @xmath5 strip ( kissr 924 ) .
the above numbers represent the luminosity contributions from the rosat - detected subsample of the kiss objects
. however , as discussed in section [ predictions ] , it is possible that those objects which are too faint to have been detected by rosat may nonetheless be significant contributors to the total volume emissivity . using x - ray luminosities estimated from the observed h@xmath12 luminosities and the adopted @xmath7 ratio
, we find that inclusion of the non - detected kiss galaxies raises the volume emissivity to @xmath89 ergs @xmath30 @xmath87 in the @xmath5 strip , and to @xmath90 ergs @xmath30 @xmath87 in the @xmath6 strip .
again , the errors are poissonian in nature , based on n = 66 agns in the @xmath5 strip and n = 36 in the @xmath6 strip .
the inclusion of objects below the rosat detection threshold increases the volume emissivity over the kiss volumes by nearly a factor of 2 , indicating that these objects do indeed have a significant effect on the global soft x - ray emission from kiss - type objects . of course , because of the prediction method employed ,
the above values represent upper limits on the volume emissivity from the kiss galaxies spectroscopically identified as agns .
one additional concern remains to be considered .
since we do not have follow - up optical spectra for all galaxies in the kiss database , selection effects are introduced into our xrb calculations .
currently , 64% and 14% of the elg sample possess follow - up spectra in the @xmath5 and @xmath6 strips , respectively .
this incompleteness results in an underestimate of the xrb contribution from the kiss elgs .
we use the properties of the @xmath5 and @xmath6 samples with follow - up spectra to correct for incompleteness effects . assuming the proportion of agns in the total galaxian population is constant
, we use the observed number of galaxies with follow - up spectra to correct for the number of additional agns expected after completion of spectral identification . after performing this correction
, we find that the total volume emissivity contributions predicted from the inclusion of non - rass detected seyfert galaxies and liners in the @xmath5 and @xmath6 strips are @xmath91 ergs @xmath30 @xmath87 and @xmath92 ergs @xmath30 @xmath87 , respectively .
the errors quoted are based on the poissonian uncertainties for the known agns in the two strips , as indicated in the preceding paragraph .
given these results for the emissivity within the survey volumes , we then calculate the x - ray intensities using the results of @xcite : @xmath93 we have ignored the possibility of redshift evolution of the x - ray luminosity function ( xlf ) in our calculations , which introduces significant uncertainties into our result .
@xcite and @xcite use population synthesis models to calculate the evolution of the soft x - ray agn luminosity function from rosat data .
the inclusion of evolutionary effects will increase the integrated x - ray intensity ( see below ) .
the fact that upper limits have been used to predict the total x - ray intensity over the kiss areas due to objects not detected by rosat will tend to compensate for the lack of consideration of evolution in our calculations , but it is not possible to determine precisely to what degree this is true . using the power - law model with spectral parameters discussed in section
[ refine ] , we then convert the 0.1 - 2.4 kev intensities from each survey strip to the 0.5 - 2 kev band .
we can then apply the result of @xcite , namely , that the total 0.5 - 2 kev xrb intensity is @xmath94 ergs @xmath30 @xmath31 @xmath95 . using this value for the xrb intensity along with our corrected volume emissivities ,
we find that the total xrb contribution in the @xmath5 strip is @xmath96% , and is @xmath97% in the @xmath6 strip .
the average of these two values is @xmath98 .
the large uncertainty we quote in this final estimate reflects the scatter between the measurements from the two survey areas , rather than the poissonian uncertainty used in the errors for the xrb contributions from the individual survey strips .
we feel that this larger value is likely to be more indicative of the actual uncertainty in our result .
we stress that the predicted xrb contribution for the types of agns detected in local samples like kiss does not take into account evolution of the agn population with lookback time .
however , @xcite found that application of the pure density evolution model derived for luminous seyfert 1s by @xcite leads to an _ overproduction _ of the 0.5 - 2 kev xrb by low - luminosity agns , given the non - evolving contribution of 9 - 11% they measured for such objects . since the xrb contribution found for the rosat - kiss sample
is somewhat higher than the halderson et al .
value , application of the miyaji evolution model will also overproduce the xrb for our sample in this energy range . while the result from this particular evolutionary model is not useful for tightening estimates of the xrb produced by low - luminosity agns , it should serve as a mild constraint on models of the evolution of x - ray luminosity from kiss - type objects .
the miyaji et al .
models are based primarily on high luminosity sources , for which the evolution with redshift may well differ from that of galaxies with lower @xmath8 .
the depth and completeness of kiss allows the survey to make unique contributions to the study of emission - line galaxies . due to its nearly volume - limited nature
, this sample will allow for an accurate determination of the statistical properties of agns .
the global properties of the kiss sample should generalize with a reasonable degree of accuracy to the low - redshift agn population as a whole .
thus we can use the ( statistically small ) sample of rosat - detected kiss agns to study the general properties of local active galaxies .
the use of the rosat all - sky survey in correlation with kiss gives rise to certain selection effects .
rosat focuses on the soft ( @xmath1 kev ) x - ray band .
thus , it is to be expected that such a survey will preferentially detect larger numbers of unabsorbed seyfert 1 galaxies .
this expectation is reflected in the rosat - kiss sample .
nearly all of the 19 seyfert 1 galaxies present in the kiss sample have been detected by the all - sky survey ; only 6 kiss sy 1 s are not .
only a very small fraction of the known kiss narrow - line agns ( seyfert 2s and liners ) are rosat - detected .
in addition , none of the kiss starburst galaxies are detected .
our estimate for the soft xrb contribution by rosat - kiss objects is roughly a factor of two higher than that found by @xcite for lower - luminosity agns with @xmath99 ergs @xmath30 .
the different values for the soft - xrb contributions found for the two studies are due , in part , to the fact that the kiss agns are on average more luminous ( both optically and in the x - rays ) than the galaxies in the halderson et al . sample .
the kiss agns are detected over a much larger volume than the very local halderson et al .
sample , meaning that the intrinsically rare luminous agns have a much better chance of showing up in kiss .
another factor is that , unlike halderson et al . who worked with a fairly complete sample of x - ray - detected agns , our calculations of the total volume emissivity and soft xrb contribution from kiss galaxies include predictions ( often upper limits ) of the x - ray luminosities of non - detected kiss objects .
the inclusion of non - detected seyfert and liner - type objects in calculations of the xrb contribution from rosat - kiss leads to a significant increase in our derived value .
the total contribution to the 0.5 - 2.0 kev xrb is approximately @xmath10 after corrections have been performed , representing the average of the values from the @xmath5 and @xmath6 strips ( @xmath100 and @xmath101 , respectively ) .
the difference between the xrb contributions from the @xmath5 and @xmath6 strips remains large ( more than a factor of 2 ) even after correcting for the incomplete nature of kiss follow - up spectral data .
this variation may be due in part to differences in large - scale structure between these two regions of the sky , but is primarily due to the presence of one or two very luminous sources in the @xmath5 strip , with no corresponding luminous objects detected in the @xmath6 strip .
our average xrb contribution represents an upper limit to the 0.5 - 2 kev x - ray xrb contribution since a significant number of the kiss objects were not detected by rosat and were therefore assigned upper - limit values for @xmath8 .
our results are consistent with recent research showing that high - redshift , more luminous agns are the dominant source of xrb emission in both the hard and soft energy bands . although objects like those in the kiss catalog can not be the dominant source of the soft xrb , it is possible that their contribution may become much more important when viewed in hard x - rays , where the effects of absorption are far less significant .
because kiss is a local sample of galaxies , it is expected that the agns in kiss will be less luminous than a similar sample at higher redshift , due to a combination of sensitivity and evolution effects .
we have not included such evolution in our calculations , a fact which may have significant impact on our results .
several authors @xcite show that the luminosity function of agns does in fact evolve significantly with redshift , and @xcite succeed in correcting for evolutionary effects when generalizing local volume emissivity to xrb contribution . however , there is some evidence that the evolution with redshift is strongest for the most luminous sources , and hence may not be directly applicable to our sample . in the absence of an evolutionary model which does not overproduce the xrb when applied to the rosat - kiss sample ,
our estimate for the soft xrb contribution is a reasonable first - order result .
we have conducted a positional correlation between the kpno international spectroscopic survey and the rosat all - sky survey databases .
there are 18 non - spurious matches in total , 72% ( 13 ) of which are spectroscopically identified as seyfert 1 objects using follow - up spectral observations .
accurate , high - resolution optical spectra are available for all rosat - kiss objects .
the fact that most soft x - ray detected kiss objects are identified as seyfert 1 galaxies is explained _ via _ selection effects .
type 1 seyferts , being less highly absorbed , are expected to be far more luminous in the soft x - ray , and their preferential detection by rosat - kiss comes as a natural consequence of the sample selection . a large proportion ( 54% ) of the seyfert 1s in the rosat - kiss sample are in fact narrow - line seyfert 1 galaxies ; this is a reflection of the high fraction of nls1s in the general kiss sample ( 44% of seyfert 1s with follow - up
spectra available to date ) .
notable non - seyfert 1 objects within the rosat - kiss sample include one seyfert 2 galaxy and two liners ; the liners are intrinsically faint objects , which may have been detected due to their proximity .
several rosat - kiss sources lie outside our adopted volume limits ; these objects have been discarded from subsequent calculations .
rosat - kiss includes one of the five qsos in the kiss database , kissr 864 .
we also note the presence of a stellar object , kissr 817 , which is bright in the soft x - ray .
kiss provides a nearly complete sample of elgs over a well - defined volume in space .
we therefore expect that the xrb contributions estimated from the rosat - kiss correlation will yield an accurate representation of the local contribution of agns to the soft xrb .
the locality of the kiss volume may imply inaccuracies in our result due to the fact that evolution has not been taken into account , but application of the density evolution model of @xcite result in overproduction of the soft xrb .
we estimate upper limits for the 0.1 - 2.4 kev @xmath8 for kiss objects not detected as x - ray sources by rosat , using the median @xmath7 from a previous study .
we also predict the number of additional agns expected to be identified after follow - up spectra have been obtained for all kiss objects , and adjust the emissivity accordingly .
we find that the upper limit for the xrb contribution from kiss agns is @xmath10 , the average over the two kiss areas .
this result is a factor of two higher than the 9 - 11% contribution estimated by @xcite for low - luminosity agns .
the differences between the two studies is primarily the depth , and hence luminosity ranges , sampled by the two surveys .
it should be remembered that the kiss - dervied xrb contribution is an upper limit , due to the manner in which the x - ray luminosities are estimated for our non - rosat - detected sources .
this project has taken a different approach to the study of the xrb . instead of studying the optical properties of an x - ray selected sample , we examine the x - ray properties of a sample of optically luminous galaxies .
future studies of the global x - ray emission properties of volume - limited galaxy samples may prove highly significant in constraining the xrb contribution from individual galaxian populations , as well as the redshift evolution of the x - ray luminosity function .
the kiss agn sample would , in fact , be an excellent one to use for establishing the x - ray characteristics for the low - redshift , low- and intermediate - luminosity agn population .
our sample size is large enough to construct an x - ray luminosity function to be used as a template for future studies of the evolution of x - ray sources .
currently , the biggest limitation is the lack of quality x - ray observations for the bulk of the kiss agn sample .
we gratefully acknowledge financial support for the kiss project from an nsf presidential faculty award to jjs ( nsf - ast-9553020 ) , as well as continued support for our ongoing follow - up spectroscopy campaign ( nsf - ast-0071114 ) .
we also thank wesleyan university for providing additional funding for several of the observing runs during which the important follow - up spectral data were obtained .
we thank the numerous kiss team members who have participated in these spectroscopic follow - up observations during the past few years , particularly caryl gronwall , drew phillips , gary wegner , laura chomiuk , jason melbourne , jeff van duyne , kerrie mckinstry , and robin ciardullo .
we appreciate help from eli beckerman and paul green regarding accessing the rosat data as well as for discussions concerning the x - ray properties of agn .
several useful suggestions by the anonymous referee helped to improve the presentation of this paper .
finally , we wish to thank the support staffs of kitt peak national observatory , lick observatory , the hobbey - eberly telescope , mdm observatory , and apache point observatory for their excellent assistance in obtaining both the survey data as well as the spectroscopic observations . without their assistance
this project would have been impossible .
bade , n. , et al .
1995 , , 110 , 469 boller , t. , bertoldi , f. , dennefeld , m. , & voges , w. 1998 , a&as , 129 , 87 boller , t. , brandt , w. n. , & fink , h. 1996 , a&a , 305 , 53 brinkmann , w. , et al .
2000 , , 356 , 445 comastri , a. , setti , g. , zamorani , g. , & hasinger , g. 1995 , , 296 , 1 dickey , j. m. & lockman , f. j. 1990 , , 28 , 215d fabian , a. c. , & barcons , x. 1992 , , 30 , 429 geller , m. j. et al .
1997 , , 114 , 2205 giacconi , r. , gursky , h. , paolini , f. r. , & rossi , b. b. 1962 , , 9 , 439 gilli , r. , salvati , m. , & hasinger , g. 2001 , , 366 , 407 gioia , i.m .
, et al . 1990 , , 72 , 567 goodrich , r. w. 1989 , , 342 , 224 griffiths , r. e. , della ceca , r. , georgantopoulos , i. , boyle , b. j. , stewart , g. c. , shanks .
t. , fruscione , a. 1996 , , 281 , 71 gronwall , c. , salzer , j. j. , sarajedini , v. l. , jangren , a. , chomiuk , l. b. , moody , j. w. , frattare , l. m. , & boroson , t. a. 2002a , in preparation ( kr2 ) gronwall , c. , salzer , j. j. , jangren , a. , wegner , g. a. , phillips , a. c. , melbourne , j. , ciardullo , r. , & mckinstry , k. 2002b , in preparation gronwall , c. et al .
2002c , in preparation grupe , d , beuermann , k. , mannheim , k. , & thomas , h .- c . 1999 , , 350 , 805 halderson , e. l. , moran , e. c. , filippenko , a. v. , & ho , l. c. 2001 , , 122 , 637 hasinger , g. , et al . 1998 , , 329 , 482 ho , l. c. , et al . 2001 , , 549 , 51 hornschemeier et al .
2001 , , 554 , 742 macalpine , g. m. , smith , s. b. , & lewis , d. w. 1977 , , 34 , 95 macalpine , g. m. , & williams , g. a. 1981 , , 45 , 113 markarian , b. e. 1967 , astrofizika , 3 , 55 markarian , b. e. , lipovetskii , v. a. , & stepanian , d. a. 1981 , astrofizika , 17 , 619 ( english transl .
astrophysics , 19 , 14 ) melbourne , j. , phillips , a. c. , salzer , j. j , gronwall , c. , & sarajedini , v. l. 2002 , in preparation miyaji , t. , hasinger , g. , & schmidt , m. 2000 , , 353 , 25 miyaji , t. , hasinger , g. , & schmidt , m. 2001 , , 369 , 49 moran , e. c. , kay , l. e. , davis , m. , filippenko , a. v. , barth , a. j. 2001 , , 556 , l75 osterbrock , d. e. , & pogge , r. w. 1985 , , 297 , 166 pesch , p. , & sanduleak , n. 1983 , , 51 , 171 rosati , p. , et al .
2002 , , 566 , 667 salzer , j. j. , et al . 2000 , , 120 , 80 salzer , j. j. , et al . 2001 , , 121 , 66 ( kr1 ) salzer , j. j. , et al . 2002 , , 123 , 1292 salzer , j. j. , gronwall , c. , jangren , a. , & mckinstry , k. 2003 , in preparation schmidt , m. , et al . 1998 , , 329 , 495 snowden , s. l. , et al .
1995 , , 454 , 643 snowden , s. l. , et al .
1997 , 485 , 125 soltan , a. m. , hasinger , g. , egger , r. , snowden , s. , truemper , j. 1996 , , 305 , 17 stephens , s. a. 1989 , , 97 , 10 stephenson , c. b. , pesch , p. , & macconnell , d. j. 1992 , , 82 , 471 tozzi , p. , et al .
2001 , , 562 , 42 ueda , y. , et al .
1999 , , 518 , 656 voges , w. , et al . 1999 , , 349 , 389 wegner , g. a. , et al .
2001 , , 122 , 2893 wegner , g. a. , salzer , j. j. , jangren , a. , gronwall , c. , & melbourne , j. 2002 , in preparation white , n. e. , giommi , p. , angelini , l. 1994 , , 6100 , 1 zamorano , j. , gallego , j. , rego , m. , vitores , a. g. , & alonso , o. 1996 , , 105 , 343 rrllrcrrrrrr 29&j122006.2 + 291643&12:20:06.8&29:16:50.3&12.1&liner&-18.06&11.53&0.99&0.00203&0.036&39.48 + 281&j132438.5 + 291020&13:24:38.7&29:10:11.9&8.6&sy 1&-19.43&17.95&0.72&0.07257&0.050&42.72 + 350&j134607.5 + 293814&13:46:08.1&29:38:10.7&9.5&sy 1&-20.97&16.51&0.47&0.07663&0.261&43.49 + 360&j135020.1 + 291005&13:50:19.7&29:10:03.2&6.3&sy 1&-18.05&19.13&1.20&0.06680&0.021&42.27 + 489&j141434.6 + 293439&14:14:34.5&29:34:28.7&10.4&sy 1&-19.61&17.85&0.50&0.07590&0.041&42.68 + 531&j142219.8 + 294257&14:22:20.2&29:42:55.5&6.1&sy 1&-20.67&15.99&0.60&0.05289&0.105&42.78 + 583&j145109.1 + 292633&14:51:09.4&29:26:26.0&8.0&sy 1&-20.19&17.30&1.22&0.07563&0.137&43.22 + 632&j150620.7 + 294005&15:06:20.8&29:39:57.3&8.4&sy 1&-19.57&17.36&0.93&0.05895&0.028&42.33 + 817&j153619.3 + 292146&15:36:18.6&29:21:50.9&11.4&cv ( ? ) & ....... & 19.34&-0.09&0.00000&0.058 & ....... + 864&j154842.8 + 290901&15:48:43.7&29:08:59.8&13.3&qso&-24.33&17.70&0.06&0.55010&0.049&44.61 + 924&j155643.0 + 294838&15:56:42.9&29:48:48.0&9.7&sy 1&-21.76&16.08&0.68&0.08554&0.457&43.91 + 1494&j131325.3 + 433219&13:13:25.8&43:32:14.2&9.6&sy 2&-20.89&15.95&0.82&0.05732&0.015&42.02 + 1503&j131406.3 + 434649&13:14:07.1&43:46:34.1&19.7&sy 1&-19.40&18.36&1.14&0.08723&0.026&42.62 + 1511&j131511.5 + 432544&13:15:10.1&43:25:46.8&21.8&sy 1&-20.16&17.58&0.62&0.08614&0.024&42.58 + 1755&j141755.1 + 431152&14:17:55.5&43:11:55.2&6.8&sy 1&-19.90&17.66&0.57&0.07902&0.201&43.41 + 1802&j143239.7 + 430021&14:32:41.1&43:00:39.8&28.5&sy 1&-21.78&19.57&0.11&0.42133&0.038&44.20 + 1903&j150506.8 + 435002&15:05:07.3&43:50:05.2&8.3&sy 1&-20.21&17.74&1.00&0.09359&0.069&43.12 + 2059&j155339.7 + 434410&15:53:40.4&43:44:04.3&11.9&liner&-19.67&16.48&0.90&0.04048&0.017&41.77 + | we present a study of the x - ray properties of a volume - limited sample of optically selected emission - line galaxies .
the sample is derived from a correlation between the kpno international spectroscopic survey ( kiss ) , an @xmath0-selected objective - prism survey of agns and starbursting galaxies , and the rosat all - sky survey ( rass ) .
after elimination of all spurious matches , we identify 18 rosat - detected x - ray sources within the kiss sample in the @xmath1 kev band . due to soft x - ray selection effects , the majority of the rosat sources are seyfert 1 galaxies . the majority ( 54% ) of the rosat - kiss seyferts
are classified as narrow - line seyfert 1 objects , a relatively high percentage compared to previous objective - prism - selected seyfert galaxy samples .
we estimate the x - ray luminosities of the rosat - detected kiss objects and derive volume emissivities of @xmath2 @xmath3 and @xmath4 @xmath3 for the @xmath5 and @xmath6 strips , respectively . for those kiss agns
not detected by rass , we use the median @xmath7 ratio derived from a previous study to estimate @xmath8 .
the total @xmath9 kev volume emissivity we predict for the overall kiss agn sample is sufficient to account for @xmath10 of the soft x - ray background ( xrb ) , averaged over both survey strips .
the kiss agn sample is made up predominantly of intermediate - luminosity seyfert 2 s and liners , which tend to be weak soft x - ray sources .
they may , however , represent a much more significant contribution to the hard xrb . |
in a previous paper @xcite we have motivated and studied the hamiltonian thermodynamics of three - dimensional dilatonic black holes . to construct a classical hamiltonian formalism
is important for several reasons .
first , it is always elegant to write the equations of the theory as a pair of symmetric first order equations .
second , if one can put the classical theory into a hamiltonian form , then by applying certain rules one can get a quantized version of the theory in a first approximation .
third , through euclideanization of the schrdinger time evolution operator @xmath12 , where @xmath13 is the hamiltonian , one gets the partition function , which in turn leads to a thermodynamical decription of the system .
thus , the hamiltonian formalism and the thermodynamical description are linked subjects .
the hamiltonian thermodynamics of several different black hole systems have been studied first by louko and whiting in @xcite and further analyzed in @xcite and @xcite , ( see also @xcite ) .
the louko - whiting method relies on the hamiltonian approach of @xcite , which in turn is an important ramification of the arnowitt , deser , and misner approach , the adm approach , and its major developments @xcite .
other methods , like direct calculation from quantum fields in curved spacetime ( see , e.g , @xcite ) , or path integral methods ( see , e.g ,
@xcite ) , have been used to study the thermodynamics of black holes .
now , the study of thermodynamics of black holes in any specific dimension is important , for instance , to understand universal properties independent of the dimension itself . in particular , one can single out three dimensions as an adequate interesting dimension , because three dimensional general relativity has a special black hole with simple properties @xcite , and dilatonic extensions of this theory have black holes with properties similar to four - dimensional general relativistic black holes @xcite .
the main purpose of this paper is to generalize our previous work on hamiltonian thermodynamics of three - dimensional dilatonic black holes @xcite , to black holes , still in three dimensions , in which the graviton - dilaton brans - dicke theory , with its @xmath0 parameter , is coupled to a gauge field , namely , the electromagnetic field .
three - dimensional einstein - maxwell theory with its black hole solution has been studied in @xcite ( see also @xcite ) , and electric dilatonic extensions also have charged black holes @xcite ( see also @xcite ) .
further analyses on the properties of these black holes have been worked out in @xcite . in principle
, @xmath0 can take any value , i.e. , @xmath14 .
but , since we want to perform a canonical hamiltonian analysis , using an adm formalism supplied with proper boundary conditions , it is necessary to pick up only those @xmath0s which yield black hole solutions that fulfill boundary conditions adequate to a hamiltonain method .
as in the uncharged case @xcite , the cases of most interest are then black holes for which @xmath15 , @xmath2 , and @xmath3 .
now , unlike the uncharged case , it was realized in @xcite ( see also @xcite ) that the inclusion of electric charge creates a complication to the definition of the mass @xmath5 .
indeed , for @xmath6 , which includes the general relativity case ( since it can be considered as the @xmath16 limit ) , one finds a divergent asymptotic behavior of the mass .
this problem can be cured by a proper mass renormalization @xcite .
but the whole issue has never been confronted when one makes a full use of the hamiltonian formalism , as we do here .
we solve this difficulty and show that the mass redefinition conforms wholly with the formalism for the @xmath17 case , is not needed in the @xmath2 case and it is compatible with the @xmath3 case . in each case
, the black hole system and its thermodynamics can then be studied .
we have chosen to study the systems in a grand canonical ensemble .
in such an ensemble , the boundary radius @xmath18 , the temperature @xmath19 ( or its inverse @xmath20 ) , and the electric potential ( i.e. , the chemical potential ) @xmath21 at the boundary radius are fixed .
the electric potential is conjugate to the electric charge and , in this ensemble , the charge itself is not fixed .
our main results are , a complete derivation from a hamiltonian formalism of the mass - energy @xmath22 , the entropy @xmath23 , and the charge @xmath24 , as well as the establishment that the grand canonical ensemble for charged three dimensional black holes is locally stable for the black holes studied .
thus , the choice of such an ensemble is appropriate , since the corresponding boundary onditions lead to a well posed problem and a stable ensemble .
the structure of the paper is as follows . in sec .
[ solucoes ] we present the classical solutions of the three - dimensional dilatonic charged black holes , whose quantization through hamiltonian methods we will perform .
there is a free parameter @xmath0 for which we choose three different values , @xmath25 , corresponding to the btz black hole , the dimensionally reduce four - dimensional cylindrical black hole , and a three - dimensional dilatonic black hole , respectively .
we also introduce the spacetime foliation through which we will define the canonical coordinates and which will allow us to write the action as a sum of constraints multiplied by their respective lagrange multipliers .
then , follow three sections , secs .
[ mtz ] , [ zero ] , and [ menostres ] , where we develop the thermodynamic hamiltonian formalism for the @xmath25 charged black holes , respectively .
we divide each section into a - f subsections to better analyze the hamiltonian thermodynamics of the systems .
three - dimensional theories which contain black holes have been studied in several works . indeed in @xcite , a brans - dicke action , with gravitational and dilaton fields plus a cosmological constant , was used to study such solutions .
now , we can couple the graviton and dilaton fields to an electromagnetic field . in order
to find charged black hole solutions in three dimensions . for three - dimensional general relativity
this has been done in @xcite ( see also @xcite ) , whereas for brans - dicke theory this has been done in @xcite ( see also @xcite ) .
the brans - dicke action with a maxwell term is given by @xmath26 where @xmath27 is the determinant of the three - dimensional metric @xmath28 , @xmath29 is the curvature scalar , @xmath30 is a scalar dilaton field , @xmath0 is the brans - dicke parameter , @xmath31 is the cosmological constant , @xmath32 is the maxwell tensor , where @xmath33 is the vector potential one - form , and @xmath34 is a generic surface term . a general solution is a solution for the metric , the dilaton and the gauge field .
we search for static solutions .
the solution for the metric is then @xcite @xmath35 dt^2 + \frac{dr^2}{(ar)^2 - \frac{b}{(ar)^{\frac{1}{\omega+1 } } } + \frac{k\chi^2}{(ar)^{\frac{2}{\omega+1}}}}+ r^2 d\varphi^2 , \,\,\,\ , \omega\neq-2\,,-\frac{3}{2}\,,-1\ , , \label{solutions_1}\\ ds^2 & = & -\left[\left(1+\frac{\chi^2}{4}\ln r \right)r^2-br \right ] dt^2 + \frac{dr^2}{\left(1+\frac{\chi^2}{4}\ln r \right)r^2-br } + r^2 d\varphi^2 \qquad \qquad \quad \,\,\,\,\ , \omega=-2\,,\\ ds^2 & = & \left(4\lambda^2 r^2 \ln(br ) + \chi^2 r^4\right ) dt^2 - \frac{dr^2}{4\lambda^2 r^2 \ln(br ) + \chi^2 r^4}+ r^2 d\varphi^2 , \qquad \qquad \qquad \qquad \ , \omega=-\frac{3}{2}\,,\\ ds^2 & = & -dt^2 + dr^2 + d\varphi^2 , \qquad \qquad \qquad \qquad \qquad \qquad \qquad \qquad \qquad \qquad \qquad \qquad \qquad \,\,\,\ , \omega=-1 , \label{mink } \end{aligned}\ ] ] where @xmath36 are schwarzschild coordinates , @xmath37 is a constant related to the cosmological constant ( see below ) , and @xmath38 is a constant of integration related to the mass ( see below ) , @xmath39 is constant of integration related to the electric charge , and @xmath40 is defined as @xmath41 .
the general solution for @xmath30 is given by @xmath42 the general solution for the vector potential @xmath43 , written as @xmath44 , is given by @xmath45 for the constant @xmath37 one has @xmath46 for @xmath47 we have @xmath48 , which means that , unlike for the uncharged case , the cosmological constant is not null .
the constant @xmath38 is related to the adm mass of the solutions by @xmath49 the constant @xmath39 is s related to the adm charge of the solutions by @xmath50 for @xmath51 , there is no good definition for @xmath9 in terms of @xmath39 .
since we want to perform a canonical hamiltonian analysis , using an adm formalism supplied with proper boundary conditions , it is necessary to pick up only those solutions that fulfill the conditions we want to impose .
first , we are interested only in solutions with horizons , so we take @xmath38 to be positive .
second , apart from a measure zero of solutions , all solutions have a non - zero @xmath52 .
this does not mean straight away that the solutions are asymptotically anti - de sitter .
some have one type or another of singularities at infinity , which do not allow an imposition of proper boundary conditions .
so , from @xcite with the corresponding carter - penrose diagrams we discard the following solutions : @xmath53 which is simply the minkowski solution , @xmath54 since it has a curvature singularity at @xmath55 , which is inside the horizon , and has a null topological singularity at @xmath56 , @xmath57 since all the carter - penrose boundary is singular , the same happening in the interval @xmath58 .
thus , the cases of interest to be studied are black holes for which @xmath59 and @xmath60 . for @xmath38
positive these solutions have adm mass @xmath5 positive . as in @xcite
we choose three typical amenable cases where an analytical study can be done .
the first case is the charged btz black hole @xcite ( see also @xcite ) , which can be found by taking appropriately the limit @xmath1 .
this black hole is thus a solution of the einstein - maxwell action in three dimensions .
the other cases are @xmath2 and @xmath3 .
the theory with @xmath2 is equivalent to cylindrical four - dimensional einstein - maxwell general relativity , and the corresponding charge black hole was found in @xcite , see also @xcite .
the theory with @xmath3 is just a case of 3d charged brans - dicke theory , with a black hole solution that can be analyzed in this context @xcite .
now , as mentioned in the introduction , the definition of the adm mass @xmath5 is not straightforward when there is electric charge , in the sense that one can face a divergent asymptotic behavior of the mass @xmath5 , see @xcite as well as @xcite .
the problem is more acute when @xmath6 , which includes the general relativity case ( since general relativity can be considered the @xmath16 ) , see @xcite for a full discussion ( see also @xcite ) .
for such @xmath0 , the mass @xmath5 can be written as @xmath61 , where @xmath62 is the mass for the uncharged case and @xmath63 is the divergent part when there is charge . to treat this divergence ,
as explained in @xcite , we consider a large radius @xmath64 frontier .
then we write the mass @xmath5 as @xmath65\ , , \label{mass_ren } \end{aligned}\ ] ] where the function @xmath66 is equal to @xmath67 the term in square brackets in eq .
( [ mass_ren ] ) tends to zero as the radius @xmath29 tends to the frontier radius , @xmath68 .
@xmath66 is interpreted as the energy inside the radius @xmath64 . from eq .
( [ mass_uncharged ] ) the term @xmath69 is interpreted as the energy outside the frontier in @xmath64 , apart from an infinite constant , which is hidden in @xmath66 . the total mass definition , @xmath5 ,
is then well defined . in practical terms
, one needs not pay attention to the divergent factors @xmath70 because all one needs to do is consider null the asymptotic limits , @xmath71 , ignoring the frontier at @xmath64 . for @xmath72
the mass term behavior of the three - dimensional black holes is similar to the reissner - nordstrm case , and it offers no problems . indeed , for the reissner - nordstrm black hole the metric function @xmath73 and the mass term , following the prescription above , are written as @xmath74 , and @xmath75 , respectively .
the energy inside the frontier @xmath64 is then given by @xmath66 , and the electrostatic energy outside is given by @xmath76 .
as this last term tends to zero as @xmath77 there are no real divergences in this case , thus implying that @xmath66 carries no infinity within , and so the prescription is trivial , it gves nothing new .
this is similar to what happens for three - dimensional black holes with @xmath72 . on the other hand , for @xmath78
one should follow the prescription carefully , i.e. , one has to hide an infinite constant in the definition of the energy within the frontier at @xmath64 .
the ansatz for the metric and dilaton fields with which we start our canonical analysis is given by @xmath79 @xmath80 where @xmath81 are the schwarzschild type time , radius , and angular coordinates .
this is the adm ansatz for spherically symmetric solutions applied to three dimensions . in this
we follow the basic formalism developed by kucha @xcite .
the canonical coordinates @xmath29 and @xmath82 are functions of @xmath83 and @xmath84 , i.e. , @xmath85 and @xmath86 .
now , @xmath87 is generically on the horizon as analyzed in @xcite , but for our purposes @xmath87 represents the horizon bifurcation point of the carter - penrose diagram @xcite ( see also @xcite-@xcite and @xcite ) . in three spacetime dimensions the point actually represents a circle .
asymptotically the coordinate @xmath84 tends to @xmath88 , and @xmath83 is another time coordinate .
the remaining functions are the lapse function @xmath89 , and the shift function @xmath90 , and they will play the role of lagrange multipliers of the hamiltonian of the theory .
the canonical coordinates @xmath91 and the lapse function @xmath89 are taken to be positive .
the angular coordinates are left untouched , due to spherical symmetry .
the dilaton is a simple function of the radial canonical coordinate , and it can be traded directly by it through equation ( [ admphiansatz ] ) , as will be done below . the ansatz ( [ adm_ansatz])-([admphiansatz ] ) is written in order to perform the foliation of spacetime into spacelike hypersurfaces , and thus separates the spatial part of the spacetime from the temporal part .
indeed , the canonical analysis requires the explicit separation of the time coordinate from the other space coordinates , and so in all expressions time is treated separately from the other coordinates .
it breaks explicit , but not implicit , covariance of the theory , but it is necessary in order to perform the hamiltonian analysis .
the metric coefficients of the induced metric on the hypersurfaces become the canonical variables , and the momenta are determined in the usual way , by replacing the time derivatives of the canonical variables , the velocities . then , using the hamiltonian , one builds a time evolution operator to construct an appropriate thermodynamic ensemble for the geometries of a quantum theory of gravity . assuming
that a quantum theory only makes sense if its classical form can be quantized by hamiltonian methods , one should pick up only solutions which can be put consistently in a hamiltonian form .
thus , in the following we perform a hamiltonian analysis to extract the entropy and other thermodynamic properties in the three - dimensional brans - dicke - maxwell black holes mentioned above , those for which @xmath92 .
the canonical description of the vector potential one - form is given by @xmath93 where @xmath94 and @xmath95 are functions of @xmath83 and @xmath84 , i. e. , @xmath96 and @xmath97 .
the function @xmath96 is the canonical coordinate associated with the electric field , and the function @xmath97 is the lagrange multiplier associated with the electromagnetic constraint , which is gauss law . in the same way as with the gravitational degrees of freedom above ,
when the maxwell term in the action ( [ 3dchbdaction ] ) is written , the conjugate momentum to the coordinate @xmath96 can be derived the usual way from the lagrangian .
for @xmath98 , the corresponding three - dimensional charged brans - dicke theory is precisely the one provided by three - dimensional einstein - maxwell theory @xcite .
then the general metric in eq .
( [ solutions ] ) , the @xmath30 field in eq .
( [ phisolutions_1 ] ) , and the vector potential one - form ( [ vector_potential ] ) reduce to the following @xmath99 where @xmath18 is the ads length , related to the cosmological constant by @xmath100 , @xmath5 is the mass , and @xmath9 is the electric charge . in this solution
the scalar field @xmath30 is trivial .
next , in fig .
[ omega_infinity_maxwell ] , we show the carter - penrose diagram of the three - dimensional charged black hole . despite there being a large and rich structure , the outside region in i is the relevant one ,
thus our foliation is from the bifurcation point to @xmath101 , and our boundary conditions reflect properties of the outer horizon at @xmath102 , and of the spacelike infinity at @xmath101 .
[ htmb ] omega_infinity_maxwell.eps given eqs .
( [ adm_ansatz])-([a_ansatz ] ) , the three - dimensional action ( [ 3dchbdaction ] ) , now an einstein - maxwell action , becomes , excluding surface terms , @xmath103 & = & \int dt \int_0^{\infty } dr \left\{- 2n^{-1}\dot{\lambda}\dot{r } + 2 n^{-1}n^r r'\dot{\lambda } + 2 n^{-1}(n^r)'\lambda \dot{r } + 2 n^{-1}n^r \lambda ' \dot{r}\right .
\nonumber \\ & & \left .
- 2n^{-1}(n^r)^2\lambda ' r ' + 2n\lambda^{-2}\lambda ' r'-2n\lambda^{-1}r '' + 4n\lambda^2\lambda r \right .
\nonumber \\ & & \left .
+ \pi n^{-1}\lambda^{-1}r\left(\dot{\gamma}-\phi'\right)^2 \right\}\ , , \label{infty } \end{aligned}\ ] ] where @xmath100 and @xmath18 is the ads length , which then yields the quantity @xmath37 is @xmath104 . in eq .
( [ infty ] ) a @xmath105 means derivative with respect to time @xmath83 and a @xmath106 means derivative with respect to @xmath84 , and where all the explicit functional dependences are omitted . from the action ( [ infty ] ) one determines the canonical momenta , conjugate to @xmath82 , @xmath29 , and @xmath94 respectively ,
@xmath107 by performing a legendre transformation , we obtain @xmath108 defining implicitly in this way the constraints @xmath13 and @xmath109 , and @xmath110 , and where the new lagrange multiplier is @xmath111 . here
additional surface terms have been ignored , as for now we are interested in the bulk terms only .
the action in hamiltonian form is then @xmath112 = \int dt \int_0^{\infty } dr \left\{p_{\lambda}\dot{\lambda}+p_r\dot{r}+p_{\gamma}\dot{\gamma } -nh - n^rh_r-\tilde{\phi}g\right\}\,.\ ] ] the equations of motion are @xmath113 in order to have a well defined variational principle , we need to eliminate the surface terms of the original bulk action , which render the original action itself ill defined when one seeks a correct determination of the equations of motion through variational methods .
these surface terms are eliminated through judicious choice of extra surface terms which should be added to the action .
the action ( [ haction_oo ] ) has the following extra surface terms , after variation @xmath114 in order to evaluate this expression , we need to know the asymptotic conditions of each of the above functions individually , which are functions of @xmath115 . starting with the limit @xmath116 we assume @xmath117
note that there are time dependences on the left hand side of the falloff conditions and that there are no such dependences on the right hand side , in the lower orders of the expansion in @xmath84 .
this apparent discrepancy stems from the fact that there is in fact no time dependence in the lower orders of the majority of the functions , but there may still exist such dependence for higher orders
. nevertheless , terms such as @xmath118 are functions , independent of @xmath115 , thus constant , but undetermined .
their variation makes sense , as we may still vary between different values for these constant functions . with these conditions
, we have for the surface terms at @xmath87 @xmath119 for @xmath120 we have @xmath121 these conditions imply for the surface terms in the limit @xmath120 @xmath122 where @xmath123 .
so , the surface term added to ( [ haction_oo ] ) is @xmath124 = \int dt \left(2 r_0 n_1 \lambda_0^{-1 } - \tilde{n}_+ m_+ + \tilde{\phi}_0 \left(2\pi q_0\right ) -
\tilde{\phi}_+ \left(2\pi q_+\right)\right)\,.\ ] ] what is left after varying this last surface term and adding it to the varied initial action ( see eq .
( [ haction_oo ] ) ) is @xmath125 we choose to fix @xmath126 and @xmath127 on the horizon ( @xmath87 ) , and @xmath128 and @xmath129 at infinity .
these choices make the surface variation ( [ variationofsurfaceterm_oo ] ) disappear .
the term @xmath126 is the integrand of @xmath130 which is the rate of the boost suffered by the future unit normal to the constant @xmath83 hypersurfaces defined at the bifurcation circle , i.e. , at @xmath131 , due to the evolution of the constant @xmath83 hypersurfaces . by fixing the integrand we are fixing the rate of the boost , which allows us to control the metric singularity when @xmath131 @xcite .
finally , we also fix @xmath129 and @xmath132 at @xmath87 and at infinity , respectively . in order to reconstruct the mass and the time from the canonical data , which amounts to making a canonical transformation
, we have to rewrite the form of the solutions of eqs .
( [ solucao_mtz])-([vector_potential_oo ] ) .
however , in addition to reconstructing the mass as done in @xcite , we have to consider how to reconstruct the charge from the canonical data .
we follow kucha @xcite for this reconstruction .
we concentrate our analysis on the right static region of the carter - penrose diagram . developing the killing time @xmath133 as function of @xmath115 in the expression for the vector potential ( [ vector_potential_oo ] ) , and making use of the gauge freedom that allows us to write @xmath134 where @xmath135 is an arbitrary continuous function of @xmath83 and @xmath84 , we can write the one - form potential as @xmath136 equating the expression ( [ vector_potential_plusgauge_tr_oo ] ) with eq .
( [ a_ansatz ] ) , and making use of the definition of @xmath137 in eq .
( [ p_gamma_oo ] ) , we arrive at @xmath138 in the static region , we define @xmath139 as @xmath140 where we define the @xmath64 as the ( large ) radius of a finite frontier , which will serve as a renormalization of the asymptotic properties of the future canonical coordinate @xmath141 ( see subsection [ bhsolutions ] ) . this frontier will be used throughout the present section .
we now make the following substitutions @xmath142 into the solution ( [ solucao_mtz ] ) , getting @xmath143 this introduces the adm foliation directly into the solutions . comparing it with the adm metric ( [ adm_ansatz ] ) , written in another form as @xmath144 we can write a set of three equations @xmath145 the first two equations , eqs .
( [ adm_1_oo ] ) and eq .
( [ adm_2_oo ] ) , give @xmath146 this one solution , together with eq .
( [ adm_1_oo ] ) , give @xmath147 one can show that @xmath148 is positive ( see @xcite ) . next , putting eqs .
( [ shift_def_oo])-([lapse_def_oo ] ) , into the definition of the conjugate momentum of the canonical coordinate @xmath82 , given in eq .
( [ p_lambda_oo ] ) , one finds the spatial derivative of @xmath149 as a function of the canonical coordinates , i.e. , @xmath150 later we will see that @xmath151 , as it will be conjugate to a new canonical coordinate @xmath5 .
following this procedure to the end , we may then find the form of the new coordinate @xmath141 , as a function of @xmath83 and @xmath84 .
first , we need to know the form of @xmath139 as a function of the canonical pair @xmath152 . for that
, we replace back into eq .
( [ adm_1_oo ] ) the definition , in eq .
( [ t_linha_oo ] ) , of @xmath153 , giving @xmath154 equating this form of @xmath139 with eq .
( [ f_1_oo ] ) , we obtain @xmath155 where @xmath139 is given in eq .
( [ f_2_oo ] ) .
we thus have found the form of the new canonical coordinate , @xmath5 .
it is now a straightforward calculation to determine the poisson bracket of this variable with @xmath156 and see that they are conjugate , thus making eq .
( [ t_linha_oo ] ) the conjugate momentum of @xmath5 , i.e. , @xmath157 the other natural transformation has been given in ( [ p2piq ] ) , namely the one that relates @xmath137 with @xmath9 through @xmath158 .
as @xmath5 , @xmath9 is also a natural choice for canonical coordinate , being another physical parameter of the solution ( [ solucao_mtz ] ) .
the one coordinate which remains to be found is @xmath10 , the conjugate momentum to the charge .
it is also necessary to find out the other new canonical variable which commutes with @xmath5 , @xmath8 , and @xmath9 , and which guarantees , with its conjugate momentum , that the transformation from @xmath159 , to @xmath5 , @xmath9 , and the new variable is canonical .
immediately it is seen that @xmath29 commutes with @xmath5 , @xmath8 , and @xmath9 .
it is then a candidate .
it remains to be seen whether @xmath160 also commutes with @xmath5 , @xmath8 , and @xmath9 . as with @xmath29 ,
it is straightforward to see that @xmath160 does not commute with @xmath5 and @xmath8 , as these contain powers of @xmath29 in their definitions , and @xmath161 .
so , rename the canonical variable @xmath29 as @xmath162 .
we have then to find a new conjugate momentum to @xmath163 which also commutes with @xmath5 , @xmath8 , and @xmath9 , making the transformation from @xmath164 a canonical one .
the way to proceed is to look at the constraint @xmath109 , which is called in this formalism the super - momentum .
this is the constraint which generates spatial diffeomorphisms in all variables .
its form , in the initial canonical coordinates , is @xmath165 . in this formulation
, @xmath82 is a spatial density , @xmath29 is a spatial scalar , and @xmath94 is also a spatial density . as the new variables , @xmath5 , @xmath163 , and @xmath9 , are spatial scalars , the generator of spatial diffeomorphisms is written as @xmath166 , regardless of the particular form of the canonical coordinate transformation .
it is thus equating these two expressions of the super - momentum @xmath109 , with @xmath5 , @xmath8 , and @xmath9 written as functions of @xmath159 and their respective momenta , that gives us one equation for the new @xmath167 and @xmath10 .
this means that we have two unknowns , @xmath167 and @xmath10 , for one equation only .
this sugests that we should make the coefficients of @xmath168 and @xmath169 equal to zero independently .
this results in @xmath170 it can be shown that @xmath10 commutes with the new @xmath167 and with the rest of the new coordinates , except with @xmath9 .
we have now all the canonical variables of the new set determined . for completeness and future use ,
we write the inverse transformation for @xmath82 and @xmath171 , @xmath172 in summary , the canonical transformations are the following , @xmath173 it remains to be seen that this set of transformations is in fact canonical . in order to prove that the set of equalities in expression ( [ setcanonicaltrans_oo ] ) is canonical we start with the equality @xmath174 we now integrate expression ( [ oddidentity_oo ] ) in @xmath84 , in the interval from @xmath87 to @xmath175 . the first term on the right hand side of eq .
( [ oddidentity_oo ] ) vanishes due to the falloff conditions ( see eqs .
( [ falloff_0_i_oo])-([falloff_0_f_oo ] ) and eqs .
( [ falloff_inf_i_oo])-([falloff_inf_f_oo ] ) ) .
we then obtain the following expression @xmath176\ , , \end{aligned}\ ] ] where @xmath177 $ ] is a well defined functional , which is also an exact form .
this equality shows that the difference between the liouville form of @xmath178 and the liouville form of @xmath179 is an exact form , which implies that the transformation of variables given by the set of equations ( [ setcanonicaltrans_oo ] ) is canonical .
armed with the certainty of the canonicity of the new variables , we can write the asymptotic form of the canonical variables and of the metric function @xmath180 . these are , for @xmath116 @xmath181 with @xmath182 for @xmath183 , we have @xmath184 where @xmath123 , as seen before in the surface terms ( see eq .
( [ surface_inf_oo ] ) ) .
note here that without the treatment of the divergence in subsection [ bhsolutions ] the function @xmath141 would diverge .
we are now almost ready to write the action with the new canonical variables .
it is now necessary to determine the new lagrange multipliers . in order to write the new constraints with the new lagrange multipliers
, we can use the identity given by the space derivative of @xmath5 , @xmath185 solving for @xmath13 and making use of the inverse transformations of @xmath82 and @xmath171 , in eqs .
( [ inversetrans_1_oo ] ) and ( [ inversetrans_2_oo ] ) , we get @xmath186 following kucha @xcite , the new set of constraints , totally equivalent to the old set @xmath187 , @xmath188 , and @xmath189 , outside the horizon points , is @xmath190 , @xmath191 , and @xmath192 . by continuity
, this also applies on the horizon , where @xmath193 .
so we can say that the equivalence is valid everywhere .
the new hamiltonian , which is the total sum of the constraints , can now be written as @xmath194 in order to determine the new lagrange multipliers , one has to write the left hand side of the previous equation , eq .
( [ newhamiltonian_oo ] ) , and replace the constraints on that side by their expressions as functions of the new canonical coordinates , spelt out in eqs .
( [ oldconstraintsinnewvariables_1_oo])-([oldconstraintsinnewvariables_2_oo ] ) .
after manipulation , one gets @xmath195 using the inverse transformations eqs .
( [ inversetrans_1_oo])-([inversetrans_2_oo ] ) , and the identity @xmath162 , with @xmath196 , we can write the new multipliers as functions of the old variables @xmath197 allowing us to determine its asymptotic conditions from the original conditions given above . these transformations are non - singular for @xmath198 . as before , for @xmath116 , @xmath199 and for @xmath120 we have @xmath200 the conditions ( [ mult_newfalloff_0_i_oo])-([mult_newfalloff_inf_f_oo ] ) show that the transformations in eqs .
( [ mult_trans_1_oo])-([mult_trans_2_oo ] ) are satisfactory in the case of @xmath120 , but not for @xmath116 .
this is due to fact that in order to fix the lagrange multipliers for @xmath120 , as we are free to do , we fix @xmath201 , which we already do when adding the surface term @xmath202 to the action , in order to obtain the equations of motion in the bulk , without surface terms . the same is true for @xmath132 .
however , at @xmath87 , we see that fixing the multiplier @xmath203 to values independent of the canonical variables is not the same as fixing @xmath204 to values independent of the canonical variables .
the same is true of the fixation of @xmath205 with respect to @xmath129 .
we need to rewrite the multipliers @xmath203 and @xmath205 for the asymptotic regime @xmath116 without affecting their behavior for @xmath120 . in order to proceed we have to make one assumption , which is that the expression given in the asymptotic condition of @xmath141 , as @xmath206 , for the term of order zero , @xmath207 , defines @xmath118 as a function of @xmath208 and @xmath209 , and @xmath118 is the horizon radius function , @xmath210 .
also , we assume that @xmath211 , where @xmath212 with these assumptions , we are working in the domain of the classical solutions . we can immediately obtain that the variation of @xmath118 is given in relation to the variations of @xmath208 and @xmath209 as @xmath213 this expression will be used when we derive the equations of motion from the new action .
we now define the new multipliers @xmath214 and @xmath215 as @xmath216^{-1}\,,\\ \tilde{n}^q & = & \tilde{n}^m 2\,g\,q_0 \ln\left(\frac{r_0}{r_*}\right ) \left(2l^{-2}r_0-\pi q_0 ^ 2 r_0^{-1}\right)^{-1 } - n^q\ , , \label{new_n_q_oo } \end{aligned}\ ] ] where @xmath217 for @xmath116 and @xmath218 for @xmath120 . the new multipliers , functions of the old multipliers @xmath203 and @xmath205 , have as their properties for @xmath131 @xmath219 and as their properties for @xmath183 @xmath220 when the constraints @xmath221 hold , @xmath222 is given by @xmath223 with this new constraint @xmath214 , fixing @xmath126 at @xmath87 or fixing @xmath222 is equivalent , there being no problems with @xmath224 , which is left as determined in eq .
( [ new_mult_2_oo ] ) . with respect to @xmath215
the same happens , i. e. , fixing the zero order term of the expansion of @xmath205 for @xmath225 is the same as fixing @xmath129 . at infinity
there were no initial problems with the definitions of both @xmath214 and @xmath215 .
the new action is now written as the sum of @xmath226 , the bulk action , and @xmath227 , the surface action , @xmath228 & = & \int \,dt\ , \int_0^\infty \ , dr \ , \left ( p_m\dot{m } + p_\textrm{r } \dot{\textrm{r } } + p_q \dot{q } \right .
\nonumber \\ & & \left .
+ \tilde{n}^qq ' - n^{\textrm{r}}p_{\textrm{r } } + \tilde{n}^m ( 1-g)\,m ' \right .
\nonumber \\ & & + \tilde{n}^m 2\,g\ , \left(2l^{-2}r_0-\pi q_0 ^ 2 r_0^{-1}\right)^{-1 } \left(m'-q'q_0\ln\left(\frac{r_0}{r_*}\right)\right ) \nonumber \\ & & + \int \ , dt \ , \left\{\left(2 r_0 \tilde{n}_0^m - \tilde{n}_+ m_+ \right ) + 2\pi\left(\tilde{\phi}_0q_0-\tilde{\phi}_+q_+\right ) \right\}\ , .
\label{newaction_oo } \end{aligned}\ ] ] the new equations of motion are now @xmath229 where we understood @xmath203 to be a function of the new constraint , defined through eq .
( [ new_n_m_oo ] ) and @xmath205 as the function of the new constraint defined through eq .
( [ new_n_q_oo ] ) . the resulting boundary terms of the variation of this new action , eq . ( [ newaction_oo ] ) , are , first , terms proportional to @xmath230 , @xmath231 , and @xmath232 on the initial and final hypersurfaces , and , second , @xmath233 to arrive at ( [ surface_varied_terms_final ] ) we have used the expression in eq .
( [ var_r_m_oo ] ) .
the action in eq .
( [ newaction_oo ] ) yields the equations of motion , eqs .
( [ new_eom_1_oo])-([new_eom_9_oo ] ) , provided that we fix the initial and final values of the new canonical variables and that we also fix the values of @xmath234 and of @xmath128 , and of @xmath129 and @xmath132 .
thanks to the redefinition of the lagrange multiplier , from @xmath203 to @xmath214 , the fixation of those quantities , @xmath234 and @xmath128 , has the same meaning it had before the canonical transformations and the redefinition of @xmath203 .
the same happens with @xmath129 and @xmath132 .
this keeping of meaning is guaranteed through the use of our gauge freedom to choose the multipliers , and at the same time not fixing the boundary variations independently of the choice of lagrange multipliers , which in turn allow us to have a well defined variational principle for the action .
we now solve the constraints in order to reduce to the true dynamical degrees of freedom .
the equations of motion ( [ new_eom_1_oo])-([new_eom_9_oo ] ) allow us to write @xmath5 and @xmath9 as independent functions of space , @xmath84 , @xmath235 the reduced action , with the constraints taken into account , is then @xmath236 = \int dt\,\,\textbf{p}_{\textbf{m } } \dot{\bf{m}}+\textbf{p}_{\textbf{q } } \dot{\bf{q}}-\textbf{h}\,,\ ] ] where @xmath237 and the reduced hamiltonian , @xmath238 , is now written as @xmath239 with @xmath240 being the horizon radius .
we also have that @xmath241 , according to the assumptions made in the previous subsection .
thanks to the functions @xmath242 , @xmath201 , @xmath243 , and @xmath244 the hamiltonian @xmath238 is an explicitly time dependent function . the variational principle associated with the reduced action , eq .
( [ red_action_oo ] ) , will fix the values of @xmath245 and @xmath246 on the initial and final hypersurfaces , or in the spirit of the classical analytical mechanics , the hamiltonian principle fixes the initial and final values of the canonical coordinates .
the equations of motion are @xmath247 the equation of motion for @xmath245 , eq .
( [ red_eom_1_oo ] ) , is understood as saying that @xmath245 is , on a classical solution , equal to the mass parameter @xmath5 of the solution , eq .
( [ solucao_mtz ] ) .
the same goes for the function @xmath246 , where eq .
( [ red_eom_2_oo ] ) implies that @xmath246 is equal to the charge parameter @xmath9 on a classical solution , eq .
( [ solucao_mtz ] ) . in order to interpret the other equation of motion , eq .
( [ red_eom_3_oo ] ) , we have to recall that from eq .
( [ p_m_oo ] ) one has @xmath156 , where @xmath133 is the killing time .
this , together with the definition of @xmath248 , given in eq .
( [ new_p_m_oo ] ) , yields @xmath249 where @xmath250 is the value of the killing time at the left end of the hypersurface of a certain @xmath83 , and @xmath251 is the killing time at spatial infinity , the right end of the same hypersurface of @xmath83 . as the hypersurface evolves in the spacetime of the black hole solution , the right hand side of eq .
( [ red_eom_2_oo ] ) is equal to @xmath252 . finally , after the definition @xmath253 obtained from eqs .
( [ vector_potential_plusgauge_tr_oo ] ) , ( [ p_q_oo ] ) , and ( [ new_p_q_oo ] ) , eq .
( [ red_eom_4_oo ] ) gives in the right hand side @xmath254 , which is the diference of the time derivatives of the electromagnetic gauge @xmath135 at @xmath87 and at infinity .
the next step is to quantize the reduced hamiltonian theory , by building the time evolution operator quantum mechanically and then obtaining a partition function through the analytic continuation of the same operator @xcite-@xcite .
the variables @xmath245 and @xmath246 are regarded here as configuration variables .
these variables satisfy the inequality @xmath241 .
the wave functions will be of the form @xmath255 , with the inner product given by @xmath256 where @xmath43 is the domain of integration defined by @xmath241 and @xmath257 is a smooth and positive weight factor for the integration measure .
it is assumed that @xmath258 is a slow varying function , otherwise arbitrary .
we are thus working in the hilbert space defined as @xmath259 .
the hamiltonian operator , written as @xmath260 , acts through pointwise multiplication by the function @xmath261 , which on a function of our working hilbert space reads @xmath262 this hamiltonian operator is an unbounded essentially self - adjoint operator .
the corresponding time evolution operator in the same hilbert space , which is unitary due to the fact that the hamiltonian operator is self - adjoint , is @xmath263\ , .
\label{k}\ ] ] this operator acts also by pointwise multiplication in the hilbert space .
we now define @xmath264 using ( [ red_hamiltonian_oo ] ) , ( [ k])-([csi_+_oo ] ) we write the function @xmath265 , which is in fact the action of the operator in the hilbert space , as @xmath266\ , .
\label{k2}\ ] ] this expression indicates that @xmath267 depends on @xmath268 and @xmath269 only through the functions @xmath270 , @xmath271 , @xmath272 , and @xmath273 .
thus , the operator corresponding to the function @xmath265 can now be written as @xmath274 .
the composition law in time @xmath275 can be regarded as a sum of the parameters @xmath276 , @xmath271 , @xmath272 , and @xmath273 inside the operator @xmath274 .
these parameters are evolutions parameters defined by the boundary conditions , i.e. , @xmath270 is the killing time elapsed at right spatial infinity and @xmath271 is the boost parameter elapsed at the bifurcation circle ; @xmath272 and @xmath273 are line integrals along timelike curves of constant @xmath84 , and constant angular variables , at @xmath87 and at infinity
. we can now build the partition function for this system .
the path to follow is to continue the operator to imaginary time and take the trace over a complete orthogonal basis .
our classical thermodynamic situation consists of a three - dimensional spherically symmetric charged black hole , asymptotically anti - de sitter , in thermal equilibrium with a bath of hawking radiation . ignoring back reaction from the radiation ,
the geometry is described by the solutions in eqs .
( [ solucao_mtz])-([vector_potential_oo ] ) .
thus , we consider a thermodynamic ensemble in which the temperature , or more appropriately here , the inverse temperature @xmath20 , and the electric potential @xmath21 are fixed .
this characterizes a grand canonical ensemble , and the partition function @xmath277 arises naturally in such an ensemble .
to analytically continue the lorentzian solution we put @xmath278 , and @xmath279 , this latter choice based on the regularity of the classical euclidean solution .
we also choose @xmath280 and @xmath281 .
we arrive then at the following expression for the partition function @xmath282\,.\ ] ] from eq .
( [ k2 ] ) this is realized as @xmath283\left\langle \textbf{m}|\textbf{m}\right\rangle\ , .
\label{z1}\ ] ] since @xmath284 is equal to @xmath285 , one has to regularize eq .
( [ z1 ] ) .
following the procedure developed in the louko - whiting approach @xcite , this means regularizing and normalizing the operator @xmath286 beforehand .
this leads to @xmath287\,,\ ] ] where @xmath288 is a normalization factor and @xmath289 is the domain of integration .
provided the weight factor @xmath258 is slowly varying compared to the exponential in eq .
( [ partition_function_3_oo ] ) , and using the fact that the horizon radius @xmath240 is a function of @xmath245 and @xmath246 , the integral in eq .
( [ partition_function_3_oo ] ) is convergent . changing integration variables , from @xmath245 to @xmath240 , keeping @xmath246 , where @xmath290 the integral eq .
( [ partition_function_3_oo ] ) becomes @xmath291 where @xmath292 is new the domain of integration after changing variables , and the function @xmath293 , a kind of an effective action ( see @xcite ) , is written as @xmath294 the new domain of integration , @xmath295 , is defined by the inequalities @xmath296 and @xmath297 .
the new weight factor @xmath298 includes the jacobian of the transformation .
since the weight factor @xmath298 is slowly varying , we can estimate the integral of @xmath299 by the saddle point approximation .
for that we have to calculate the critical points
. firstly one finds the value of @xmath246 for which @xmath300 the value is @xmath301^{-1}\ , .
\end{aligned}\ ] ] replacing this value in ( [ eff_action_oo ] ) , we obtain @xmath302^{-1}\right ) -4\pi r_{\textrm{h}}\ , .
\end{aligned}\ ] ] deriving expression ( [ eff_action_at_q_star_oo ] ) with respect to @xmath240 and making it zero , we obtain @xmath303^{-2 } r_{\textrm{h}}^{-1}-4\pi=0\ , , \end{aligned}\ ] ] which implies @xmath304^{-2 } -4\pi r_{\textrm{h}}&=&0\ , . \end{aligned}\ ] ] the solutions @xmath305 to the equation ( [ solution_poly_oo ] ) gives us the pair @xmath306 of critical points , with @xmath307 given in eq .
( [ q_star ] ) , where @xmath308 . of these
, the classical solution comes from the one critical pair that is also the global minimum of the effective action @xmath309 , denoted by @xmath310 , depending on the interval of values to which @xmath21 belongs .
the renormalized partition function is then @xmath311\,,\ ] ] where @xmath312 is a slowly varying prefactor and @xmath310 is the global minimum of the effective action ( [ eff_action_oo ] ) . in the domain of integration
the dominating contribution comes from the vicinity of @xmath313 .
we now write the logarithm of @xmath314 as @xmath315 by ignoring the prefactor s logarithm , which closer to @xmath316 is less relevant , we are able to determine the value of @xmath245 at the critical point , where we find that it corresponds to the value of the mass of the classical solutions of the black holes ( see eq .
( [ solucao_mtz ] ) ) .
thus , when the critical point dominates the partition function , we have that the mean energy @xmath22 is given by @xmath317 where @xmath318 is obtained from eq .
( [ m_r_h_oo ] ) evaluated at @xmath316 .
the thermal expectation value of the charge is @xmath319 the temperature of the black hole , @xmath320 , is @xmath321 where @xmath322 is the function in ( [ q_star ] ) evaluated at @xmath316 .
if the maximum value of the charge @xmath322 is chosen , i.e. , @xmath323 , which is the value of the charge of the extreme solution of the black hole , and is replaced into ( [ temperature_oo ] ) , then the temperature is null , as expected from an extreme solution .
it can be shown that @xmath324 , which through the constant @xmath21 heat capacity @xmath325 tells us that the system is thermodynamically stable .
the entropy is given by @xmath326 this is the entropy of the charged btz black hole @xcite ( see also @xcite ) .
this entropy includes the extreme solution , as the specific value of the charge is irrelevant for the determination of the entropy , provided we are in the domain of validity of the approximation and @xmath327 .
[ htmb ] omega_0.eps for @xmath2 , the corresponding three - dimensional charged brans - dicke theory is obtained from the cylindrical dimensionally reduced black hole of four - dimensional general relativity , with a maxwell term @xcite .
then general metric in eq .
( [ solutions ] ) , the @xmath30 field in eq .
( [ phisolutions_1 ] ) , and the vector potential one - form ( [ vector_potential ] ) reduce to the following @xmath328\,dt^2 + \frac{dr^2}{(a\,r)^2 - \frac{m}{2(a\,r)}+ \frac{q^2}{16(a\,r)^2}}+r^2\,d\varphi^2\ , , \label{metric_zero}\\ e^{-2\phi } & = & a\,r\,,\label{phi_zero}\\ a & = & \alpha^{-1}\frac{q}{r}\,dt \ , , \label{vector_potential_zero } \end{aligned}\ ] ] with @xmath329 , @xmath330 , and @xmath331 , where the @xmath18 is the ads length .
the schwarzschild coordinates are again @xmath81 . unlike the charged btz solution
, this solution ( [ metric_zero])-([vector_potential_zero ] ) has a metric function whose mass term depends on @xmath29 .
this behavior is similar to the reissner - nrdstrom black hole metric function @xcite . in fig .
[ omega_0 ] we show the carter - penrose diagram of the black hole solution for @xmath2 , where again @xmath240 is the horizon radius , given by the larger positive real root of @xmath332 @xmath56 is the radius of the timelike curvature singularity , and @xmath101 is the spatial infinity .
the action with @xmath2 becomes , following eqs .
( [ adm_ansatz])-([a_ansatz ] ) , and excluding surface terms , @xmath103 & = & \int dt \int_0^{\infty } dr \,\alpha \left\{\lambda^2\lambda n r^2 -n^{-1}r\dot{r}\dot{\lambda}-\frac12 n^{-1}\lambda\dot{r}^2+n^{-1 } r\dot{r } ( \lambda n^r)'\right .
\nonumber\\ & & + n^{-1}n^r r r'\dot{\lambda } - \left . n^{-1}n^r\lambda r r'(n^r ) ' -n^{-1}(n^r)^2rr'\lambda'+n^{-1}n^r\lambda\dot{r}r ' \right .
\nonumber\\ & & \left .
-\frac12 n^{-1}(n^r)^2\lambda(r')^2 -(\lambda^{-1})'rr'n -\frac12 \lambda^{-1}(r')^2n-\lambda^{-1}rr''n\right . \nonumber \\ & & \left .
+ \frac12 \alpha n^{-1}\lambda^{-1}r^2\left(\dot{\gamma}- \phi'\right)^2\right\}\ , , \end{aligned}\ ] ] where @xmath105 means derivative with respect to time @xmath83 and @xmath106 is the derivative with respect to @xmath84 , and where all the explicit functional dependences are omitted .
depending on the situation we use four different letters containing the same information but with slightly different numerical values .
thus , @xmath37 , @xmath31 , @xmath18 and @xmath333 are related by @xmath334 , and @xmath331 , where @xmath18 is the ads length . from this action
one determines the canonical momenta , conjugate to @xmath82 , @xmath29 , and @xmath94 respectively @xmath335+\lambda[\dot{r}-n^rr ' ] \right\}\ , , \label{p_r}\\ p_\gamma & = & \alpha n^{-1}\lambda^{-1}r^2\left(\dot{\gamma}- \phi'\right)\ , .
\label{p_gamma } \end{aligned}\ ] ] by performing a legendre transformation , we obtain @xmath336 with the new lagrange multiplier being @xmath337 . the action in hamiltonian form is then @xmath338 = \int dt \int_0^{\infty } dr \left\ { p_{\lambda}\dot{\lambda}+p_r\dot{r}+p_\gamma\dot{\gamma } -nh - n^rh_r-\tilde{\phi}g \right\}\,.\ ] ] the equations of motion are @xmath339 in order to have a well defined variational principle , we need to eliminate the surface terms of the original bulk action , which render the original action itself ill defined for a correct determination of the equations of motion through variational methods . through the choice of added surface terms one can achieve this elimination . the action ( [ haction ] ) has the following extra surface terms , after variation @xmath340 in order to evaluate this expression , we need to know the asymptotic conditions of each of the functions of @xmath115 .
starting with the limit @xmath116 we assume @xmath341 with these conditions , we have for the surface terms at @xmath87 , @xmath342 in the same way , for @xmath120 , @xmath343 these conditions imply for the surface terms in the limit @xmath120 , @xmath344 where @xmath345 .
so , the surface term to be added to ( [ haction ] ) is @xmath346= \int dt \left(\frac12 \alpha r_0 ^ 2 n_1 \lambda_0^{-1 } - \tilde{n}_+ m_+ + \tilde{\phi}_0 q_0 - \tilde{\phi}_+ q_+ \right)\,.\ ] ] what is left after varying this last surface term and adding it to the varied initial action ( see eq .
( [ haction ] ) ) is @xmath347 we choose to fix @xmath126 and @xmath129 on the horizon , and @xmath128 and @xmath132 at infinity , which makes the surface variation ( [ variationofsurfaceterm ] ) disappear . in order to reconstruct the mass and the time from the canonical data , which amounts to making a canonical transformation
, we have to rewrite the general form of the solutions in eqs .
( [ metric_zero])-([vector_potential_zero ] )
. however , in addition to reconstructing the mass as done in @xcite , we have to consider how to reconstruct the charge from the canonical data .
we follow kucha @xcite for this reconstruction .
we concentrate our analysis on the right static region of the carter - penrose diagram . developing the killing time @xmath133 as function of @xmath115 in the expression for the vector potential ( [ vector_potential_zero ] ) , and making use of the gauge freedom that allows us to write @xmath348 where @xmath135 is an arbitrary continuous function of @xmath83 and @xmath84 , we can write the one - form potential as @xmath349 equating the expression ( [ vector_potential_plusgauge_tr_zero ] ) with eq .
( [ a_ansatz ] ) , and making use of the definition of @xmath137 in eq .
( [ p_gamma ] ) , we arrive at @xmath350 in the right static region we define @xmath139 as @xmath351 note that here , @xmath64 , the large radius of a finite frontier , which can be used to renormalize the asymptotic properties of the future canonical coordinate @xmath141 ( see subsections [ bhsolutions ] and [ reconstruction ] ) , is not needed ; in fact bringing to the @xmath2 solution introduces complications at the @xmath87 frontier .
thus , in this section we do not mention @xmath64 . then , making the following substitutions @xmath142 in the solutions above , eqs .
( [ metric_zero])-([phi_zero ] ) , one has @xmath143 this introduces the adm foliation directly into the solutions . comparing it with the adm metric ( [ adm_ansatz ] ) , written in another form as @xmath144 we can write a set of three equations @xmath352 the first two equations , eq .
( [ adm_1 ] ) and eq .
( [ adm_2 ] ) , give @xmath353 this one solution , together with eq .
( [ adm_1 ] ) , give @xmath354 one can show that @xmath148 is positive .
next , putting eq .
( [ shift_def ] ) and eq .
( [ lapse_def ] ) into the definition of the conjugate momentum of the canonical coordinate @xmath82 , given in eq .
( [ p_lambda ] ) , one finds the spatial derivative of @xmath149 as a function of the canonical coordinates , i.e. , @xmath355 later we will see that @xmath151 , as it will be conjugate to a new canonical coordinate @xmath5 .
following this procedure to the end , we may then find the form of the new coordinate @xmath141 , also as a function of @xmath83 and @xmath84 .
first , we need to know the form of @xmath139 as a function of the canonical pair @xmath152 . for that
, we replace back into eq .
( [ adm_1 ] ) the definition of @xmath153 , giving @xmath356 equating this form of @xmath139 with eq .
( [ f_1 ] ) , we obtain @xmath357 where @xmath139 is given in eq .
( [ f_2 ] ) .
we thus have found the form of the new canonical coordinate , @xmath5 .
it is now a straightforward calculation to determine the poisson bracket of this variable with @xmath156 and see that they are conjugate , thus making eq .
( [ t_linha ] ) the conjugate momentum of @xmath5 , i.e. , @xmath358 the other natural transformation has been given , namely the one that relates @xmath137 with @xmath9 through @xmath359 .
as @xmath5 , @xmath9 is also a natural choice for canonical coordinate , being another physical parameter of the solution ( [ solucao_mtz ] ) .
the one coordinate which remains to be found is @xmath10 , the conjugate momentum to the charge .
it is also necessary to find out the other new canonical variable which commutes with @xmath5 , @xmath8 , and @xmath9 , and which guarantees , with its conjugate momentum , that the transformation from @xmath159 , to @xmath5 , @xmath9 , and the new variable is canonical .
immediately is it seen that @xmath29 commutes with @xmath5 , @xmath8 , and @xmath9 .
it is then a candidate .
it remains to be seen whether @xmath160 also commutes with @xmath5 , @xmath8 , and @xmath9 . as with @xmath29 ,
it is straightforward to see that @xmath160 does not commute with @xmath5 and @xmath8 , as these contain powers of @xmath29 in their definitions , and @xmath161 .
so rename the canonical variable @xmath29 as @xmath162 .
we have then to find a new conjugate momentum to @xmath163 which also commutes with @xmath5 , @xmath8 , and @xmath9 , making the transformation from @xmath360 a canonical one .
the way to proceed is to look at the constraint @xmath109 , which is called in this formalism the super - momentum .
this is the constraint which generates spatial diffeomorphisms in all variables .
its form , in the initial canonical coordinates , is @xmath165 . in this formulation
, @xmath82 is a spatial density , @xmath29 is a spatial scalar , and @xmath94 is also a spatial density . as the new variables , @xmath5 , @xmath163 , and @xmath9 , are spatial scalars , the generator of spatial diffeomorphisms is written as @xmath166 , regardless of the particular form of the canonical coordinate transformation .
it is thus equating these two expressions of the super - momentum @xmath109 , with @xmath5 , @xmath8 , and @xmath9 written as functions of @xmath159 and their respective momenta , that gives us one equation for the new @xmath167 and @xmath10 .
this means that we have two unknowns , sc .
@xmath167 and @xmath10 , for one equation only .
this sugests that we should make the coefficients of @xmath168 and @xmath169 equal to zero independently .
this results in @xmath361 we have now all the canonical variables of the new set determined . for completeness and future use , we write the inverse transformation for @xmath82 and @xmath171 , @xmath362 in summary , the canonical transformations are @xmath363 in order to prove that the set of equalities in expression ( [ setcanonicaltrans ] ) is canonical we start with the equality @xmath364 we now integrate expression ( [ oddidentity ] ) in @xmath84 , in the interval from @xmath87 to @xmath175 . the first term on the right hand side of eq .
( [ oddidentity ] ) vanishes due to the falloff conditions ( see eqs .
( [ falloff_0_i])-([falloff_0_f ] ) and eqs .
( [ falloff_inf_i])-([falloff_inf_f ] ) ) .
we then obtain the following expression @xmath365\ , , \end{aligned}\ ] ] where @xmath366 $ ] is a well defined functional , which is also an exact form .
this equality shows that the difference between the liouville form of @xmath178 and the liouville form of @xmath179 is an exact form , which implies that the transformation of variables given by the set of equations ( [ setcanonicaltrans ] ) is canonical .
armed with the certainty of the canonicity of the new variables , we can write the asymptotic form of the canonical variables and of the metric function @xmath180 . these are , for @xmath116 @xmath367 with @xmath368 for @xmath183 , we have @xmath369 where @xmath345 , as seen before in the surface terms ( see eq .
( [ surface_inf ] ) ) .
we are now almost ready to write the action with the new canonical variables .
it is now necessary to determine the new lagrange multipliers . in order to write the new constraints with the new lagrange multipliers
, we can use the identity given by the space derivative of @xmath5 , @xmath370 solving for @xmath13 and making use of the inverse transformations of @xmath82 and @xmath171 , in eqs .
( [ inversetrans_1 ] ) and ( [ inversetrans_2 ] ) , we get @xmath371 following kucha @xcite , the new set of constraints , totally equivalent to the old set @xmath187 , @xmath188 , and @xmath372 outside the horizon , is @xmath190 , @xmath191 , and @xmath192 . by continuity , this also applies on the horizon , where @xmath193 .
so we can say that the equivalence is valid everywhere .
so , the new hamiltonian , the total sum of the constraints , can now be written as @xmath373 in order to determine the new lagrange multipliers , one has to write the left hand side of the previous equation , eq .
( [ newhamiltonian ] ) , and replace the constraints on that side by their expressions as functions of the new canonical coordinates , spelt out in eqs .
( [ oldconstraintsinnewvariables_1 ] ) -([oldconstraintsinnewvariables_3 ] ) . after manipulation
, one gets @xmath374 using the inverse transformations eqs .
( [ inversetrans_1])-([inversetrans_2 ] ) , and the identity @xmath162 , we can write the new multipliers as functions of the old variables @xmath375 allowing us to determine its asymptotic conditions from the original conditions given above . these transformations are non - singular for @xmath198 .
as before , for @xmath116 , @xmath376 and for @xmath120 we have @xmath377 these conditions ( [ mult_newfalloff_0_i])-([mult_newfalloff_inf_f ] ) show that the transformations in eqs .
( [ mult_trans_1])-([mult_trans_2 ] ) are satisfactory in the case of @xmath120 , but not for @xmath116 .
this is due to fact that in order to fix the lagrange multipliers for @xmath120 , as we are free to do , we fix @xmath201 , which we already do when adding the surface term @xmath202 to the action , in order to obtain the equations of motion in the bulk , without surface terms . the same is true for @xmath132 .
however , at @xmath87 , we see that fixing the multiplier @xmath203 to values independent of the canonical variables is not the same as fixing @xmath204 to values independent of the canonical variables .
the same is true of the fixation of @xmath205 with respect to @xmath129 .
we need to rewrite the multipliers @xmath203 and @xmath205 for the asymptotic regime @xmath116 without affecting their behavior for @xmath120 . in order to proceed we have to make one assumption , which is that the expression given in asymptotic condition of @xmath141 , as @xmath206 , for the term of order zero , @xmath378 , defines @xmath118 as a function of @xmath208 and @xmath209 , where @xmath118 is the horizon radius function , @xmath210 .
also , we assume that @xmath379 . with these assumptions
, we are working in the domain of the classical solutions .
we can immediately obtain that the variation of @xmath118 is given in relation to the variations of @xmath208 and of @xmath209 as @xmath380 this expression will be used when we derive the equations of motion from the new action .
we now define the new multipliers @xmath214 and @xmath215 as @xmath381^{-1}\ , , \\ \tilde{n}^q & = & \tilde{n}^m\,g\,q_0 \left(\frac{3\alpha^3}{32 } r_0 ^ 2 - \frac{q_0 ^ 2}{2\alpha r_0 ^ 2}\right)^{-1}-n^q\ , , \label{new_n_q } \end{aligned}\ ] ] where @xmath217 for @xmath116 and @xmath218 for @xmath120 .
these new multipliers , functions of the old multipliers @xmath203 and @xmath205 , have as their properties for @xmath183 @xmath382 and for @xmath131 @xmath383 where @xmath222 is given by @xmath384 when the constraints @xmath221 hold , the last expression is @xmath223 with this new constraint @xmath214 , fixing @xmath126 at @xmath87 or fixing @xmath234 is equivalent .
the same happens with fixing @xmath129 or @xmath215 at @xmath225 .
there are no problems with @xmath224 , which is left as determined in eq .
( [ new_mult_2 ] ) . the new action is now written as the sum of @xmath226 , the bulk action , and @xmath227 , the surface action , @xmath385 & = & \int dt\,\int_0^\infty dr \ , \left\ { p_m\dot{m } + p_\textrm{r } \dot{\textrm{r}}+p_q\dot{q } + \tilde{n}^qq'-n^{\textrm{r}}p_{\textrm{r}}\right . \nonumber\\ & + & \left.\tilde{n}^m \left[(1-g)m'+g \left(\frac{3\alpha^2}{16}r_0 ^ 2 -\frac{q_0 ^ 2}{2\alpha r_0 ^ 2}\right)^{-1 } \left(\alpha r_0m'-q_0q'\right)\right]\right\ } \nonumber \\ & + & \int dt \ , \left(\frac12 \alpha r_0 ^ 2 \tilde{n}_0^m - \tilde{n}_+ m_+ + \tilde{\phi}_0q_0-\tilde{\phi}_+q_+ \right)\ , .
\label{newaction } \end{aligned}\ ] ] the new equations of motion are now @xmath386 here we understood @xmath203 to be a function of the new constraint @xmath214 , defined through eq .
( [ new_n_m ] ) , and @xmath205 to be a function of the new constraint @xmath215 , defined through eq .
( [ new_n_q ] ) . the resulting boundary terms of the variation of this new action , eq .
( [ newaction ] ) , are , first , terms proportional to @xmath230 , @xmath231 , and @xmath232 on the initial and final hypersurfaces , and , second , the term @xmath387 here we have used the expression in eq .
( [ var_r_m ] ) .
the action in eq .
( [ newaction ] ) yields the equations of motion , eqs .
( [ new_eom_1])-([new_eom_9 ] ) , provided that we fix the initial and final values of the new canonical variables and that we also fix the values of @xmath234 , @xmath128 , @xmath129 , and @xmath132 .
thanks to the redefinition of both the lagrange multipliers , from @xmath203 to @xmath214 and from @xmath205 to @xmath215 , the fixation of those quantities , @xmath234 , @xmath128 , @xmath129 , and @xmath132 has the same meaning it had before the canonical transformations and the redefinition of @xmath203 and @xmath205 . this same meaning is guaranteed through the use of our gauge freedom to choose the multipliers , and at the same time not fixing the boundary variations independently of the choice of lagrange multipliers , which in turn allow us to have a well defined variational principle for the action .
we now solve the constraints in order to reduce to the true dynamical degrees of freedom .
the equations of motion ( [ new_eom_1])-([new_eom_9 ] ) allow us to write @xmath5 and @xmath9 as independent functions of the radial coordinate @xmath84 , @xmath388 the reduced action , with the constraints and eqs .
( [ m_t])-([q_t ] ) taken into account , is @xmath389 = \int dt\,\,\textbf{p}_{\textbf{m}}\dot{\bf{m}}+ \textbf{p}_{\textbf{q}}\dot{\bf{q } } -\textbf{h}\,,\ ] ] where @xmath390 and the reduced hamiltonian , @xmath238 , is now written as @xmath391 with @xmath240 being the horizon radius .
we also have that @xmath392 , according to the assumptions made in the previous subsection .
thanks to the functions @xmath242 , @xmath201 , @xmath243 , and @xmath244 the hamiltonian @xmath238 is an explicitly time dependent function . the variational principle associated with the reduced action , eq .
( [ red_action ] ) , will fix the value of @xmath245 and @xmath246 on the initial and final hypersurfaces , or in the spirit of the classical analytical mechanics , the hamiltonian principle fixes the initial and final values of the canonical coordinate .
the equations of motion are @xmath393 the equation of motion for @xmath245 , eq .
( [ red_eom_1 ] ) , is understood as saying that @xmath245 is , on a classical solution , equal to the mass parameter @xmath5 of the solutions in eq .
( [ metric_zero ] ) .
the same holds for @xmath246 in eq .
( [ red_eom_2 ] ) , i.e. , @xmath246 is the charge parameter in eqs .
( [ metric_zero])-([vector_potential_zero ] ) . in order to interpret equation of motion for @xmath248 , eq .
( [ red_eom_3 ] ) , we have to recall that from eq .
( [ p_m ] ) one has @xmath156 , where @xmath133 is the killing time .
this , together with the definition of @xmath248 , given in eq .
( [ new_p_m ] ) , yields @xmath249 where @xmath250 is the value of the killing time at the left end of the hypersurface of a certain @xmath83 , and @xmath251 is the killing time at spatial infinity , the right end of the same hypersurface of @xmath83 . as the hypersurface evolves in the spacetime of the black hole solutions , the right hand side of eq .
( [ red_eom_2 ] ) is equal to @xmath394 . finally , after the definition @xmath253 obtained from to eqs .
( [ vector_potential_plusgauge_tr_zero ] ) , ( [ p_q ] ) , and ( [ new_p_q ] ) , eq . ( [ red_eom_4 ] )
gives in the right hand side @xmath254 , which is the diference of the time derivatives of the electromagnetic gauge @xmath135 at @xmath87 and at infinity .
following sec .
[ mtz ] we need not repeat here the steps taken there in order to build the time evolution operator . thus the operator @xmath265 can now be written as @xmath395\ , .
\label{kappa_function_zero}\ ] ] the same composition law as in sec .
[ mtz ] holds here as well . here
again @xmath270 is the killing time elapsed at right spatial infinity and @xmath271 is the boost parameter elapsed at the bifurcation circle ; @xmath272 and @xmath273 are line integrals along timelike curves of constant @xmath84 and constant angular variables , at @xmath87 and at infinity .
we can now build the partition function for this system .
the path to follow is to continue the operator to imaginary time and take the trace over a complete orthogonal basis .
our classical thermodynamic situation consists of a three - dimensional spherically symmetric charged black hole , asymptotically anti - de sitter , in thermal equilibrium with a bath of hawking radiation . ignoring back reaction from the radiation , the geometry
is described by the solutions in eq .
( [ metric_zero ] ) and ( [ vector_potential_zero ] ) .
thus , we consider a thermodynamic ensemble in which the temperature , or more appropriately here , the inverse temperature @xmath20 is fixed .
this characterizes a grand canonical ensemble , and the partition function @xmath396 arises naturally in such an ensemble .
to analytically continue the lorentzian solution we put @xmath278 , and @xmath279 , this latter choice based on the regularity of the classical euclidean solution .
we also choose @xmath280 and @xmath397 .
we arrive then at the following expression for the partition function @xmath398\,.\ ] ] from eq .
( [ kappa_function_zero ] ) this is realized as @xmath399\left\langle \textbf{m}|\textbf{m } \right\rangle\,.\ ] ] since @xmath284 is equal to @xmath400 , one has to regularize ( [ partition_function_2 ] ) .
again , following the louko - whiting procedure @xcite-@xcite , we have to regularize and normalize the operator @xmath286 beforehand .
this leads to @xmath401\,,\ ] ] where @xmath288 is a normalization factor and @xmath43 is the domain of integration .
provided the weight factor @xmath258 is slowly varying compared to the exponential in eq .
( [ partition_function_3 ] ) , and using the fact that the horizon radius @xmath240 is function of @xmath245 and @xmath246 , the integral in eq .
( [ partition_function_3 ] ) is convergent . changing integration variables , from @xmath245 to @xmath240 , where @xmath402 the integral eq .
( [ partition_function_3 ] ) becomes @xmath403 where @xmath295 is the new domain of integration after changing variables , and the function @xmath404 , a kind of an effective action ( see @xcite ) , is written as @xmath405 the domain of integration , @xmath295 , is defined by the inequalities @xmath406 and @xmath407 .
the new weight factor @xmath298 includes the jacobian of the change of variables .
since the weight factor is slowly varying , we can estimate the integral of @xmath408 by the saddle point approximation .
the critical points are given as a pair of values @xmath409 @xmath410 the critical points belong to the domain of the effective action ( [ eff_action ] ) , and so the effective action has a null derivative at these points in eqs .
( [ crit_r_zero])-([crit_q_zero ] ) . as @xmath411
there are always two critical points , given by ( [ crit_q_zero])-([crit_r_zero ] ) .
the critical point at @xmath412 is outside the domain of physical interest as @xmath413 .
the critical point at @xmath310 is an global minimum .
the expression ( [ crit_q_zero ] ) implies that the critical points are on the straight line defined by itself , on the plane of the arguments of the effective action ( [ eff_action ] ) .
then , if we replace the value of the charge as a function of @xmath240 , i. e. , @xmath414 as given in ( [ crit_q_zero ] ) into the effective action we obtain a one variable function of @xmath240 , written as @xmath415 which is a function of @xmath240 on the straight line defined by @xmath416 .
the zeros of this effective action on the straight line are given by @xmath417 here @xmath413 is outside the physical domain of interest . as @xmath418
there are always three zeros of the function ( [ eff_action_r_zero ] ) , which means that there are no phase transitions , as the zeros of the effective action maintain their respective hierarchy whatever the value of @xmath11 , which is the order parameter of the system .
so the classical solution is found in the global minimum of the critical points @xmath310 of the effective action @xmath404 ( eq . ( [ eff_action ] ) ) .
the partition function @xmath408 can now be written , through the saddle - point approximation @xmath419\,,\\ \ln\mathcal{z}_{\textrm{ren}}(\beta,\bar{\phi } ) & \approx & \ln\textrm{p } -i_*(r_{\textrm{h}}^+,\textbf{q}^+)\ , .
\label{log_z_approx } \end{aligned}\ ] ] where @xmath312 is a slowly varying prefactor , and @xmath420 is the effective action evaluated at the global minimum critical point . by ignoring the prefactor s logarithm , which closer to @xmath310 is less relevant , we are able to determine the value of @xmath245 at the critical point , where we find that it corresponds to the value of the mass of the classical solution of the black hole given in eq .
( [ metric_zero ] ) .
thus , when the critical point dominates the partition function , we have that the mean energy @xmath421 is given by @xmath422 where @xmath318 is obtained from eq .
( [ m_r_h ] ) evaluated at @xmath310 .
the thermal expectation value of the charge is @xmath423 as given in eq .
( [ crit_q_zero ] ) .
we now can write the temperature @xmath424 as a function of the critical point given by @xmath316 and @xmath322 @xmath425 with @xmath316 explicitly written in ( [ crit_r_zero ] ) , and where @xmath322 is given in eq . ( [ crit_q_zero ] ) .
if we choose the maximum charge of the domain given by @xmath407 , then we find ourselves with the extreme solution of the black hole , whose temperature , if @xmath426 is replaced into ( [ temp_zero ] ) , is equal to zero .
it can be shown that @xmath324 , which through the constant @xmath11 heat capacity @xmath427 tells us that the system is thermodynamically stable . finally there is the entropy of the black hole , @xmath23 , which is given by @xmath428 here we see that despite the fact that the extreme solution has null temperature , it has an entropy given in ( [ entropy_zero ] ) , which is independent of the value of the charge in the domain of validity of the approximation .
for @xmath3 , the general metric in eq .
( [ solutions ] ) , the @xmath30 field in eq .
( [ phisolutions_1 ] ) , and the vector potential ( [ vector_potential ] ) reduce to the following , respectively , @xmath429\,dt^2
+ \frac{dr^2}{(a\,r)^2 - 2(a\,r)^\frac12 \,m-\frac12(a\,r)q^2 } + r^2\,d\varphi^2\ , , \label{metric_dilatonic}\\ e^{-2\phi } & = & \frac{1}{(a\,r)^\frac12}\ , , \label{phi_dilatonic}\\ a & = & -\frac12(a\,r)^{\frac12}q dt\ , , \label{vector_potential_dilatonic } \end{aligned}\ ] ] with @xmath430 and @xmath431 . [ htmb ] omega_-3.eps in fig .
[ omega_-3 ] the carter - penrose diagram of the black hole solution for the case @xmath3 is shown , which is analogous to the case @xmath2 , where again the singularity at @xmath56 is a curvature singularity , and @xmath101 is spatial infinity .
the horizon radius is given by the largest real solution of the following equation @xmath432 note that this diagram , fig .
[ omega_-3 ] , holds for static black holes with positive mass @xmath433 . here
i and i are the right and left static outer regions of the carter - penrose diagram , respectively , and ii and ii are the future and past inner dynamical regions of the carter - penrose diagram .
@xmath434 is the horizon radius given in eq .
( [ horizon_radius_dilatonic ] ) and the spacelike infinity is asymptotically anti - de sitter , both in the right and in the left regions , denoted by i and i , respectively .
there are no extreme black hole solutions for @xmath3 @xcite .
since we are now familiar with the whole formalism , we will be briefer in this section omitting several of the details .
the action becomes , following eqs .
( [ adm_ansatz])-([a_ansatz ] ) , and up to surface terms , @xmath435 & = & \int dt \int_0^{\infty } dr\ , \left\{4\,(a\,r)^{-\frac12}n\lambda^2\lambda r- ( a\,r)^{-\frac12}n^{-1}\dot{\lambda}\dot{r}+ \frac14 ( a\,r)^{-\frac12 } n^{-1 } \lambda r^{-1 } \dot{r}^2\right .
\nonumber \\ & & + ( a\,r)^{-\frac12 } n^{-1 } \dot{r } ( n^r \lambda ) ' + ( a\,r)^{-\frac12 } n^{-1}n^r\dot{\lambda}r ' -(a\,r)^{-\frac12 } n^{-1}n^r(n^r)'\lambda r ' \nonumber \\ & & -(a\,r)^{-\frac12 } n^{-1}(n^r)^2 \lambda ' r ' -\frac12 ( a\,r)^{-\frac12 } n^{-1}n^r \lambda r^{-1}\dot{r}r ' + \frac14 ( a\,r)^{-\frac12 } n^{-1}(n^r)^2\lambda r^{-1}(r')^2 \nonumber \\ & & \left
. + \frac14 ( a\,r)^{-\frac12 } n \lambda^{-1}r^{-1}(r')^2 - ( a\,r)^{-\frac12 } n ( \lambda^{-1})'r ' -(a\,r)^{-\frac12}n\lambda^{-1}r''\right
. \nonumber \\ & & + \left .
2\,n^{-1}\lambda^{-1}(a\,r)^{-\frac12}r \left(\dot{\gamma}-\phi'\right)^2\right\}\ , .
\end{aligned}\ ] ] depending on the situation we use three different letters containing the same information but with slightly different numerical values .
thus , @xmath37 , @xmath31 , and @xmath18 are related by @xmath436 , where @xmath18 is defined now as the ads length . from the action
above , we obtain the conjugate momenta @xmath437 by performing a legendre transformation we obtain the hamiltonian , which is a sum of constraints , i.e. , @xmath438 \right . \nonumber \\ & & \left .
-4(a\,r)^{-\frac12}\lambda^2\lambda r + \frac18 n\lambda r^{-1}(a\,r)^{\frac12}p_\gamma^2 \right\ } \nonumber \\ & & + \ , n^r\left\{p_r r ' - \lambda p_\lambda ' -\gamma p_\gamma ' \right\ } + \tilde{\phi}\left\{-p_\gamma'\right\}\ , \equiv\,nh+n^r h_r+\tilde{\phi}g\ , .
\end{aligned}\ ] ] we can now write the action in hamiltonian form , which reads @xmath439 = \int dt \int_0^{\infty } dr \left\ { p_{\lambda}\dot{\lambda}+p_r\dot{r } + p_\gamma\dot{\gamma}-nh - n^rh_r-\tilde{\phi}g \right\}\,,\ ] ] with the constraints defined in eq . ( [ hamiltonian_-3 ] ) .
from here we derive the equations of motion for the canonical variables and respective canonical momenta @xmath440 for a correct variational principle to be applied , we have to find out what surface terms are left over from the variation performed , with the purpose of deriving the equations of motion in eqs .
( [ eom_-3_1])-([eom_-3_6 ] ) .
these surface terms are , @xmath441 in order to know the form of ( [ surf_terms_-3 ] ) for @xmath442 , we assume @xmath443 the surface terms become @xmath444 the asymptotic conditions for @xmath120 are assumed as @xmath445 where @xmath18 is the ads length .
the surface terms , eq .
( [ surf_terms_-3 ] ) , for @xmath120 , are written as @xmath446 where @xmath447 here we pay attention to the fact that we can define the above quantity @xmath448 , in eq .
( [ m_+_-3 ] ) , because we do as explained in subsection [ bhsolutions ] when @xmath6 and absorb the divergence in the definition of @xmath5 .
this means that the @xmath448 above in eq .
( [ m_+_-3 ] ) is in fact related to the mass plus the divergence , defined above as @xmath66 , see eq .
( [ mass_uncharged ] ) and below in subsection [ rec_cantrans_action_-3 ] .
the surface term added to ( [ haction_-3 ] ) is then @xmath449= \int dt \left(2a^{-1}(a\,r_0)^{\frac12}n_1 \lambda_0^{-1 } -\tilde{n}_+m_++\tilde{\phi}_0q_0-\tilde{\phi}_+q_+\right)\,.\ ] ] with this surface term added we obtain a well defined variational principle . what remains after variation of the total action , eq .
( [ haction_-3 ] ) and eq .
( [ variationofsurfaceterm_-3 ] ) , is @xmath450 the surface terms coming from the variation of the total action disappear as the result of the fixation of @xmath451 on the horizon , @xmath87 , and of @xmath201 at infinity , @xmath452 , with the same happening respectively to @xmath129 at @xmath87 and to @xmath132 at @xmath453 . repeating the steps of the two previous sections , we now give the main results concerning @xmath3 .
the vector potential is @xmath454 where there is the function @xmath455 .
expanding for @xmath115 we find @xmath456 out of which we infer that @xmath457 the metric function is given by the expression @xmath458 the killing time @xmath459 is a function of @xmath115 , where we find that @xmath460 which is equal to minus the conjugate momentum of the new variable @xmath5 , or @xmath461 . with the help of eq .
( [ tlinha_-3 ] ) we find @xmath462 as a function of the canonical variables , @xmath463 summing up , the canonical transformations are @xmath464 \nonumber \\ & & + \frac14 r^{-1}\lambda p_\lambda - 4\lambda^2f^{-1}p_\lambda r\lambda+\frac18\,a\,f^{-1}\lambda p_\lambda p_\gamma^2 \ , , \nonumber \\
p_m & = & f^{-1 } \lambda p_\lambda ( a\,r)^{\frac12}\ , , \nonumber \\
p_q & = & -\gamma + \frac12 ( a\,r)\,\lambda p_\lambda p_\gamma\ , .
\end{aligned}\ ] ] in addition , the relevant inverse transformations back to the old variable @xmath82 and respective conjugate momentum @xmath171 are @xmath465 we have to show that this transformation is canonical .
this requires using the identity @xmath466 we now integrate eq .
( [ oddidentity_-3 ] ) , in @xmath84 , from @xmath87 to @xmath175 . the first term on the right hand side of eq .
( [ oddidentity_-3 ] ) vanishes due to the falloff conditions , eqs .
( [ falloff_0_i_-3])-([falloff_0_f_-3 ] ) and eqs .
( [ falloff_inf_i_-3])-([falloff_inf_f_-3 ] ) .
we obtain then the following expression @xmath467\ , , \end{aligned}\ ] ] where @xmath366 $ ] is a well defined functional , which is also an exact form . as
above , this equality shows that the difference between the liouville form of @xmath178 and the liouville form of @xmath179 is an exact form , which implies that the set of transformations ( [ setcanonicaltrans_-3 ] ) is canonical . with this result , we write the asymptotic conditions of the new canonical coordinates for @xmath116 @xmath468 where we have @xmath469r_2 } { 4r_0(a\,r_0)^\frac12}+\frac{2a^2\,r_0 ^ 2r_2 ^ 2-\frac12 \left[2(a\,r_0)q_0q_2+a\,q_0 ^ 2r_2\right]-4r_2 ^ 2\lambda_0^{-2 } } { 2(a\,r_0)^\frac12}\,,\label{m_2_-3 } \end{aligned}\ ] ] and @xmath470 . for @xmath183 , we have @xmath471 where @xmath472 is defined in eq .
( [ m_+_-3 ] ) .
we now write the future constraint @xmath473 as a function of the older constraints @xmath474 using the inverse transformations of @xmath82 and @xmath171 in eqs .
( [ inversetrans_l_-3 ] ) and ( [ inversetrans_pl_-3 ] ) , we obtain the same form for the old constraints as functions of the new variables @xmath475 the new hamiltonian , the total sum of the constraints , can now be written as @xmath476 the new multipliers are , using eqs .
( [ oldconstraintsinnewvariables_1_-3])-([newhamiltonian_-3 ] ) , @xmath477 using the inverse transformations eqs .
( [ inversetrans_l_-3])-([inversetrans_pl_-3 ] ) , and the identity @xmath162 , we can write the new multipliers as functions of the old variables @xmath478 for @xmath116 we have , @xmath479 and for @xmath120 we have @xmath480 again , for @xmath131 , fixing @xmath481 , which means fixing @xmath482 , is not equivalent to fixing @xmath126 .
it is thus necessary to rewrite @xmath203 for @xmath131 .
the same is true of fixing @xmath483 when @xmath225 . in this case
it is not the same if one fixes @xmath484 or just @xmath129 .
so , assuming @xmath208 as a function of @xmath118 and @xmath209 allows one to define the horizon radius @xmath485 .
we are thus working in the domain where @xmath486 , the domain of the classical black hole solution .
the variation of @xmath118 is given in terms of the variations of @xmath208 and of @xmath209 in the expression @xmath487 the new multiplier @xmath214 is obtained from the old @xmath203 as @xmath488^{-1}\ , , \\
\tilde{n}^q & = & -\frac12\tilde{n}^m\,g\,q_0 \left(\frac34\,a\,(a\,r_0)^{\frac12}- \frac18a\,(a\,r_0)^{-\frac12}q_0 ^ 2\right)^{-1}-n^q\ , , \label{new_n_q_-3 } \end{aligned}\ ] ] where @xmath489 , and @xmath217 for @xmath131 and @xmath218 for @xmath120 .
this new multiplier , function of the old multiplier , @xmath214 , has as its properties for @xmath120 @xmath490 and as its properties for @xmath131 @xmath491 where @xmath222 is given by @xmath492 with this new constraint @xmath214 , fixing @xmath126 or fixing @xmath234 is equivalent , and the same happens with @xmath215 , where fixing @xmath129 is fixing the multiplier @xmath215 in the limit @xmath225 .
there are no problems with @xmath224 , which is left as determined in eq .
( [ mult_trans_2_-3 ] ) .
the new action is then , summing both the bulk and the surface terms , @xmath493 & = & \int \,dt\ , \int_0^\infty \ , dr \ ,
\left\{p_m\dot{m}+p_\textrm{r}\dot{\textrm{r}}+p_q\dot{q}+ n^qq'-n^{\textrm{r}}p_{\textrm{r}}\right.\nonumber \\ & & \left.+\tilde{n}^m \left[(1-g)m'+g\left(\frac34\,a\,(a\,r)^{\frac12}- \frac18a\,(a\,r_0)^{-\frac12}q_0 ^ 2\right)^{-1}\right.\right .
\nonumber \\ & & \left.\left.\left[(a\,r_0)^{-\frac12}m ' + q_0q'\right]\right]\right\}+ \nonumber \\ & & \int \ , dt \ , \left(2a^{-1 } ( a\,r_0)^{\frac12 } \tilde{n}_0^m - \tilde{n}_+ m_+ + \tilde{\phi}_0q_0- \tilde{\phi}_+q_+\right)\ , .
\label{newaction_-3 } \end{aligned}\ ] ] the new equations of motion are now @xmath494 here we understood @xmath203 to be a function of the new constraint , defined through eq .
( [ new_n_m_-3 ] ) . the resulting boundary terms of the variation of this new action , eq . ( [ newaction_-3 ] ) , are , first , terms proportional to @xmath230 , @xmath231 and @xmath232 on the initial and final hypersurfaces , and , second , the term @xmath495 . here
we used the expression in eq .
( [ deltar_0_-3 ] ) . the action in eq .
( [ newaction_-3 ] ) yields the equations of motion , eqs .
( [ new_eom_1_-3])-([new_eom_9_-3 ] ) , provided that we fix the initial and final values of the new canonical variables and that we also fix both the values of @xmath234 and @xmath128 , and @xmath129 and @xmath132 .
thanks to the redefinition of the lagrange multiplier , from @xmath203 to @xmath214 , and from @xmath205 to @xmath215 the fixation of those quantities , @xmath234 and @xmath128 , and @xmath129 and @xmath132 , has the same meaning it had before the canonical transformations and the redefinition of @xmath203 and @xmath205 .
this same meaning is guaranteed through the use of our gauge freedom to choose the multipliers , and at the same time not fixing the boundary variations independently of the choice of lagrange multipliers , which in turn allow us to have a well defined variational principle for the action .
we now solve the constraints in order to reduce to the true dynamical degrees of freedom .
the equations of motion , eqs .
( [ new_eom_1_-3])-([new_eom_9_-3 ] ) , allow us to write @xmath5 and @xmath9 as independent functions of the radial coordinate @xmath84 @xmath496 the reduced action , with the constraints and eq ( [ m_-3 ] ) taken into account , is @xmath497 = \int dt\,\ , \textbf{p}_{\textbf{m}}\dot{\bf{m}}+\textbf{p}_{\textbf{q}}\dot{\bf{q } } -\textbf{h}\,,\ ] ] where @xmath498 and the reduced hamiltonian , @xmath238 , is now written as @xmath499 with @xmath240 being the horizon radius .
we also have that @xmath500 .
the equations of motion are then @xmath501 here @xmath245 and @xmath246 are equal to the mass parameter @xmath5 and the charge parameter @xmath9 of the classical solution in eq .
( [ metric_dilatonic ] ) . the third equation of motion , eq .
( [ red_eom_3_-3 ] ) , describes the time evolution of the difference of the killing times on the horizon and at infinity , due to @xmath502 and eq .
( [ new_p_m_-3 ] ) . the last equation , eq .
( [ red_eom_4_-3 ] ) , gives the time evolution of @xmath503 as the difference between the time derivatives of the electromagnetic gauges @xmath135 on the horizon and at infinity due to @xmath504 and eq .
( [ new_p_q_-3 ] ) .
the steps developed in sections [ mtz ] and [ zero ] can be readily used here .
so we do not spell out the corresponding calculations in detail .
thus , obtained as above , the time evolution operator is @xmath505\ , .
\label{k_op_-3}\ ] ] we can now develop the thermodynamic analysis as in sections [ mtz ] and [ zero ] . we can now build the partition function for this system , with the hamiltonian given in eq .
( [ red_hamiltonian_-3 ] ) .
the steps are the same as was the case with @xmath98 and @xmath2 .
the thermodynamic ensemble is also the grand canonical ensemble .
again , with @xmath278 , @xmath506 , @xmath280 , and @xmath397 , we write the general form of the partition function as @xmath507\,.\ ] ] this is realized as @xmath508 where @xmath288 is a normalization factor , @xmath295 is the domain of integration given by @xmath406 and @xmath509 , and the function @xmath404 , a kind of an effective action @xcite .
@xmath404 , is written as @xmath510 its critical points are at @xmath511 given that @xmath512 there are always two critical points , but only @xmath316 is physical as it is positive , @xmath513 .
the other critical point is negative , @xmath413 .
determining the zeros of the action ( [ eff_action_-3 ] ) when eq .
( [ crit_q_-3 ] ) holds , we get @xmath514 now , for different values of @xmath515 we may have different signs for the action ( [ eff_action_-3 ] ) evaluated at the physical critical point @xmath316 .
so , for @xmath516 there are no zeros and @xmath517 , where we take @xmath518 to be a function of the critical point radius @xmath316 .
if @xmath519 then the effective action ( [ eff_action_-3 ] ) has one zero only , at @xmath316 , i. e. , @xmath520 . finally , if @xmath521 then we have two zeros , given in eq . ( [ zeros_eff_action_-3 ] ) , and @xmath522
then , despite the variation of the sign of the effective action at the physical critical point @xmath310 , there are no new minima when the order parameter @xmath11 is changed across the border given by @xmath519 , which means that there are no phase transitions , as the classical phase corresponds always to the global minimum .
we can now derive the basic thermodynamic results , as long as the approximation for the saddle point is valid , which means that we have to work in the neighborhood of the classical solution ( [ metric_dilatonic ] ) , where the critical point dominates .
the expected value of the energy @xmath523 is @xmath524 the charge expectation value is @xmath525 here we have the temperature @xmath424 given by @xmath526 it can be shown that the derivative of @xmath527 with respect to @xmath20 is negative , which means that the heat capacity is positive , i. e. , @xmath528 .
the system is thus stable .
finally the entropy is given by @xmath529 the result recovers the entropy for the three - dimensional charged dilatonic black hole with @xmath3 ( see also @xcite ) .
we have extended the work done in @xcite by adding a maxwell term to the three - dimensional brans - dicke action , and thus , have continued louko and collaborators program @xcite ( see also @xcite ) ) of studying the thermodynamic properties through hamiltonian methods of black holes in several theories in different dimensions .
specifically we have calculated the thermodynamic properties of black hole solutions with asymptotics that allow a well formulated hamiltonian formalism in three - dimensional brans - dicke - maxwell gravity .
only certain values of the brans - dicke parameter @xmath0 are allowed in this juncture .
the corresponding theories are general relativity , i.e. , @xmath1 , a dimensionally reduced cylindrical charged four - dimensional general relativity theory , i.e. , @xmath2 , and a theory representing a class of charged dilaton - gravity theories , with a typical @xmath0 given by @xmath3 , all of these coupled to a maxwell field .
we have shown that for these theories , with well defined asymptotics for geometric and electromagnetic variables , the formalism fits well .
it was possible to have a well defined electric field for all the values of @xmath0 studied .
as in @xcite several modifications were needed .
first , in the powers of the fall - off conditions , and second due to the presence of the scalar field , which was reflected in the fact that it changed the powers of the radial @xmath29 coordinate , to name a few .
of course , the addition of electromagnetism also introduced a new pair of canonical variables , which added a good deal of complexity to the calculations .
namely , the treatment of the mass term when electromagnetic interaction is included is not trivial at all in this hamiltonian formalism in three dimensions , since there is a divergent asymptotic behavior of the mass @xmath5 . within the hamiltonian formalism we solved this problem by introducing a renormalized mass .
this redefinition of the mass fits perfectly into the formalism for the @xmath530 ( i.e. , general relativity ) , it is not needed in the @xmath2 case , and conforms to the @xmath3 case .
although , in order to build a three - dimensional lorentzian hamiltonian theory with electric charge , these modifications had to be made , a quantum theory and a statistical description of the systems in question in a grand canonical ensemble were derived , and the corresponding thermodynamics , with precise values for the entropy of the black holes studied , was established .
gasd is supported by grant sfrh / bd/2003 from fct .
this work was partially funded by fundao para a cincia e tecnologia ( fct ) - portugal , through project poci / fp/63943/2005 .
j. louko and b. f. whiting , phys .
d * 51 * , 5583 ( 1995 ) .
s. bose , j. louko , l. parker , and y. peleg , phys .
d * 53 * , 5708 ( 1996 ) .
j. louko and s. n. winters - hilt , phys .
d * 54 * , 2647 ( 1996 ) .
j. louko , j. z. simon , and s. n. winters - hilt , phys .
d * 55 * 3525 ( 1997 ) .
d. r. brill , j. louko , and p. peldan , phys .
d * 56 * 3600 ( 1997 ) .
s. bose , l. parker , and y. peleg , phys .
d * 56 * , 987 ( 1997 ) .
g. kunstatter , r. petryk , and s. shelemy , phys .
d * 57 * , 3537 ( 1998 ) .
a. j. m. medved and g. kunstatter , phys .
d * 59 * , 104005 ( 1999 ) .
k. v. kucha , phys . rev .
d * 50 * , 3961 ( 1994 ) .
r. arnowitt , s. deser , and c. w. misner , in _ gravitation : an introduction to current research _ , edited by l. witten ( wiley , new york , 1962 ) , chap
. 7 [ gr - qc/0405109 ] .
t. regge and c. teitelboim , ann
. phys . * 88 * , 286 ( 1974 ) .
s. w. hawking , commun .
* 43 * , 199 ( 1975 ) .
j. d. bekenstein , phys .
d * 9 * , 3292 ( 1974 )
. s. w. hawking , in _ general relativity : an einstein centenary survey _ , edited by s.
w. hawking and w. israel ( cambridge university press , cambridge , england , 1979 ) .
j. w. york , phys .
d * 33 * , 2092 ( 1986 ) .
h. w. braden , j. d. brown , b. f. whiting , and j. w. york , phys . rev .
d * 42 * , 3376 ( 1990 ) .
o. b. zaslavskii , phys .
a * 152 * , 463 ( 1991 ) . c. s. pea and j. p. s. lemos , phys .
d * 59 * , 124007 ( 1999 ) . | the action for a class of three - dimensional dilaton - gravity theories , with an electromagnetic maxwell field and a cosmological constant , can be recast in a brans - dicke - maxwell type action , with its free @xmath0 parameter . for a negative cosmological constant ,
these theories have static , electrically charged , and spherically symmetric black hole solutions .
those theories with well formulated asymptotics are studied through a hamiltonian formalism , and their thermodynamical properties are found out .
the theories studied are general relativity ( @xmath1 ) , a dimensionally reduced cylindrical four - dimensional general relativity theory ( @xmath2 ) , and a theory representing a class of theories ( @xmath3 ) , all with a maxwell term .
the hamiltonian formalism is setup in three dimensions through foliations on the right region of the carter - penrose diagram , with the bifurcation 1-sphere as the left boundary , and anti - de sitter infinity as the right boundary .
the metric functions on the foliated hypersurfaces and the radial component of the vector potential one - form are the canonical coordinates .
the hamiltonian action is written , the hamiltonian being a sum of constraints .
one finds a new action which yields an unconstrained theory with two pairs of canonical coordinates @xmath4 , where @xmath5 is the mass parameter , which for @xmath6 and for @xmath7 needs a careful renormalization , @xmath8 is the conjugate momenta of @xmath5 , @xmath9 is the charge parameter , and @xmath10 is its conjugate momentum .
the resulting hamiltonian is a sum of boundary terms only .
a quantization of the theory is performed .
the schrdinger evolution operator is constructed , the trace is taken , and the partition function of the grand canonical ensemble is obtained , where the chemical potential is the scalar electric potential @xmath11 . like the uncharged cases studied previously , the charged black hole entropies differ , in general , from the usual quarter of the horizon area due to the dilaton . |
let @xmath2 be a hilbert space .
a set of elements @xmath3 in @xmath2 ( counting multiplicity ) is called a _ frame _ if there exist two positive constants @xmath4 and @xmath5 such that for any @xmath6 we have @xmath7 the constants @xmath4 and @xmath5 are called the _ lower frame bound _ and the _ upper frame bound _ , respectively . a frame
is called a _ tight frame _ if @xmath8 . in this paper
we focus mostly on real finite dimensional hilbert spaces with @xmath9 and @xmath10 , although we shall also discuss the extendability of the results to the complex case .
let @xmath11 $ ] .
it is called the _ frame matrix _ for @xmath12 .
it is well known that @xmath12 is a frame if and only if the @xmath13 matrix @xmath14 has rank @xmath15 .
furthermore , the optimal frames bounds are given by @xmath16 where @xmath17 are the singular values of @xmath14 . throughout this paper
we shall identify without loss of generality a frame @xmath12 by its frame matrix .
the main focus of the paper is on the _ erasure robustness _ property for a frame .
this property arise in applications such as communication where data can be lost or corrupted in the process of transmission .
suppose that we have a frame @xmath14 that is _ full spark _ in the sense that every @xmath15 columns of @xmath14 span @xmath18 , it is theoretically possible to erase up to @xmath19 data from the full set of data @xmath20 while still reconstruct the signal @xmath21 .
this is a simple consequence of the property that with the remaining available data @xmath22 with @xmath23 , @xmath21 is uniquely determined because @xmath24 . in practice , however , the condition number of the matrix @xmath25_{j\in s}$ ] could be so poor that the reconstruction is numerically unstable against the presence of additive noise in the data .
thus robustness against data loss and erasures is a highly desirable property for a frame .
there have been a number of studies that aim to address this important issue . among the first studies of erasure - robust frames
was given in @xcite .
it was shown in subsequent studies that that unit norm tight frames are optimally robust against one erasure @xcite while grassmannian frames are optimally robust against two erasures @xcite .
the literature on erasure robustness for frames is quite extensive , see e.g. also @xcite . in general , the robustness of a frame @xmath14 against @xmath26-erasures , where @xmath27 , is measured by the maximum of the condition numbers of all @xmath28 submatrices of @xmath14 .
more precisely , let @xmath29 and let @xmath30 denote the @xmath31 submatrix of @xmath14 with columns @xmath32 for @xmath33 ( in its natural order , although the order of the columns is irrelevant ) .
then the robustness against @xmath26-erasures of @xmath14 is measured by @xmath34 of course , the smaller @xmath35 is the more robust @xmath14 is against @xmath26-erasures . in @xcite , fickus and
mixon coined the term _ numerically erasure robust frame ( nerf)_. a frame @xmath14 is @xmath36-nerf if @xmath37 thus in this case @xmath38 .
note that for any full spark @xmath13 frame matrix @xmath14 and any @xmath39 there always exist @xmath40 such that @xmath14 is @xmath36-nerf .
the main goal is to find classes of frames where the bounds @xmath41 , and more importantly , @xmath42 , are independent of the dimension @xmath15 while allowing the proportion of erasures @xmath43 as large as possible .
the authors studied in @xcite the erasure robustness of @xmath44 , where the entries of @xmath45 are independent random variables of the standard normal @xmath46 distribution .
it was shown that with high probability such a matrix can be good nerfs provided that @xmath47 is no less than approximately @xmath48 of @xmath49 .
the authors also proved that equiangular frame @xmath14 in @xmath50 with @xmath51 vectors is a good nerf against up to about @xmath52% erasures .
as far as the proportion of erasures is concerned this was the best known result for nerfs .
however , the frame requires almost @xmath53 vectors .
the authors posed as an open question whether there exist nerfs with @xmath54 .
a more recent paper @xcite explored a deterministic construction based on certain group theoretic techniques .
the approach offers more flexibility in the frame design than the far more restrictive equiangular frames . in this paper
we revisit the robustness of random frames .
we provide a much stronger result for random frames , showing that for any @xmath55 , with very high probability , the frame @xmath44 is a @xmath56-nerf where @xmath57 depend only on @xmath58 and the aspect ratio @xmath59 .
one version of our result is given by the following theorem .
[ theo-1.1 ] let @xmath44 where @xmath45 is @xmath13 whose entries are independent gaussian random variables of @xmath60 distribution .
let @xmath61 .
then for any @xmath62 and @xmath63 there exist @xmath64 depending only on @xmath65 and @xmath66 such that for any @xmath67 , the frame @xmath14 is a @xmath68-nerf with probability at least @xmath69 .
later in the paper we shall provide more implicit estimates for @xmath57 that will allow us to easily compute them numerically .
note that our result is essentially the best possible , as we can not go to @xmath70 .
a corollary of the theorem is that for random gaussian frames the proportion of erasures @xmath43 can be made arbitrary large while the frames still maintain robustness with overwhelming probability .
our theorem depends crucially on a refined estimate on the smallest singular value of a random gaussian matrix .
there is a wealth of literature on random matrices .
the study of singular values of random matrices has been particularly intense in recent years due to their applications in compressive sensing for the construction of matrices with the so - called _ restricted isometry property _
( see e.g.@xcite ) .
random matrices have also been employed for phase retrieval @xcite , which aims to reconstruct a signal from the magnitudes of its samples . for a very informative and comprehensive survey of the subject
we refer the readers to @xcite , which also contains an extensive list of references ( among the notable ones @xcite ) . for the @xmath13 gaussian random matrix @xmath45 the expected value of @xmath71 and @xmath72 are asymptotically @xmath73 and @xmath74 , respectively .
many important results , such as the nerf analysis of random matrices in @xcite as well as results on the restricted isometry property in compressive sensing , often utilize known estimates of @xmath71 and @xmath72 based on hoeffding - type inequalities . one good such estimate is @xmath75 see @xcite .
the problem with this estimate is that even by taking @xmath76 we only get a bound of @xmath77 even though the probability in this case is 0 .
thus estimates such as ( [ 1.2 ] ) that cap the decay rate are often inadequate . when applied to the erasure robustness problem for frames they usually put a cap on the proportion of erasures . to go further we must prove an estimate that will allow the exponent of decay to be much larger .
we achieve this goal by proving the following theorem : [ theo-1.2 ] let @xmath45 be @xmath13 whose entries are independent random variables of standard normal @xmath60 distribution .
let @xmath61 . then for any @xmath78 there exist constants @xmath79 depending only on @xmath80 and @xmath81 such that @xmath82 furthermore
, we may take @xmath83 and @xmath84 where @xmath85 * acknowledgement .
* the author would like to thank radu balan and dustin mixon for very helpful discussions .
we begin with estimates on the extremal singular values of a ranodm matrix @xmath45 whose entries are independent standard normal random variables
. we shall assume throughout the section that @xmath45 is @xmath13 where @xmath86 .
one of the very important estimates is @xmath87 see @xcite .
our main goal of this section is to prove the estimates for smallest singular value @xmath72 stated in theorem [ theo-1.2 ] . an equivalent formulation of ( [ 2.1 ] ) is @xmath88 observe that @xmath89 where @xmath90 denotes the unit sphere in @xmath18 .
[ lem-2.1 ] let @xmath91 . for any @xmath92
the probability @xmath93 is independent of the choice of @xmath21 .
we have @xmath94 for any @xmath55 .
* proof .
* the fact that @xmath93 is independent of the choice of @xmath21 is a well know fact , which stems from the fact that the entries of @xmath95 are again independent standard normal random variables for any orthogonal @xmath96 matrix @xmath97 . in particular
, one can always find an orthogonal @xmath97 such that @xmath98 .
thus we may without loss of generality take @xmath99 . in this case @xmath100
where @xmath101 $ ] denotes the first row of @xmath45 .
. then @xmath103 has the @xmath104 distribution , which has the density function @xmath105 denote @xmath106 .
it follows that @xmath107 note that @xmath108 by stirling s formula .
the theorem now follows from @xmath109 and @xmath110 . ' '' '' a ubiquitous tool in the study of random matrices is an @xmath111-net .
f or any @xmath112 an @xmath111-net for @xmath90 is a set in @xmath90 such that any point on @xmath90 is no more than @xmath111 distance away from the set . the following result is known and can be found in @xcite : [ lem-2.2 ] for any @xmath112 there exists an @xmath111-net @xmath113 in @xmath90 with cardinality no larger than @xmath114 .
* proof of theorem [ theo-1.2 ] .
* assume that @xmath115 .
then there exists a @xmath116 such that @xmath117 .
let @xmath113 be an @xmath111-net for @xmath90 and take @xmath118 that is the closest to @xmath119 .
so @xmath120 .
thus @xmath121 hence @xmath122 note that @xmath123 by lemma [ lem-2.1 ] the first term on the right hand side is bounded from above by @xmath124 by ( [ 2.2 ] ) the second term on the right hand side is bounded from above by @xmath125 thus combining these two upper bounds we obtain the estimate @xmath126 we would like to bound @xmath127 by @xmath128 .
all we need then is to choose @xmath129 so that both upper bound terms in ( [ 2.6 ] ) are bounded by @xmath130 .
note that @xmath131 .
hence we only need @xmath132 the equation ( [ 2.8 ] ) leads to the condition @xmath133 to meet condition ( [ 2.7 ] ) we set @xmath134
. then @xmath135 .
thus ( [ 2.7 ] ) becomes @xmath136 clearly , once we fix @xmath137 and @xmath138 , say , take @xmath139 and @xmath140 , @xmath141 will be greater than the right hand side of ( [ 2.10 ] ) for small enough @xmath111 because of the condition @xmath142 .
both @xmath143 only depend on @xmath81 and @xmath80 .
the existence part of the theorem is thus proved . while we have already a good explicit estimate @xmath139 , it remains to establish the explicit formula for @xmath144 .
for any fixed @xmath138 the largest @xmath111 is achieved when ( [ 2.10 ] ) is an equality , namely @xmath145 which one can rewrite as @xmath146 where @xmath147 and @xmath148 .
it follows that @xmath149 where @xmath150 .
note that @xmath151 .
now we can take @xmath144 to be the supreme value of @xmath152 , which yields @xmath153 finally , ( [ 1.3 ] ) follows from @xmath154 and ( [ 2.2 ] ) .
the proof of the theorem is now complete . ' '' ''
* remark . *
although there does not seem to exist an explicit formula for @xmath144 given in ( [ 2.11 ] ) , there is a very good explicit approximation of it . in general
, the @xmath155 that maximize @xmath156 is rather small .
so we may approximate @xmath157 simply by @xmath158 and find the maximum of @xmath159 the maximum of @xmath160 is obtained at @xmath161 .
this @xmath162 is very close to the actual @xmath155 that maximizes @xmath156 .
thus @xmath163 has @xmath164 and it is a close approximation of the optimal @xmath144 . of course , theorem [ theo-1.2 ] still holds when @xmath144 is replaced by @xmath165 .
although theorem [ theo-1.2 ] is for real gaussian random matrices , a complex version of it can also be proved with minor modifications .
a complex random variable @xmath166 has the _ complex standard normal _
distribution if both @xmath167 and @xmath168 have the real complex normal distribution @xmath46 .
theorem [ theo-1.2 ] extends to the following theorem for the complex case : [ theo-2.3 ] let @xmath45 be @xmath13 whose entries are independent random variables of complex standard normal @xmath60 distribution .
let @xmath61 .
then for any @xmath78 there exist constants @xmath79 depending only on @xmath80 and @xmath81 such that @xmath169 furthermore , we may take @xmath170 and @xmath84 where @xmath171 * proof .
* the proof follows the same argument as in the real case so we only sketch the proof here .
in particular we point out the places where the estimates need to be modified .
write @xmath172 and set @xmath173 $ ]
. then @xmath174 is an @xmath175 matrix whose entries are independent real standard normal random variables .
it is easy to check that @xmath176 .
thus by taking @xmath177 we have via ( [ 2.1 ] ) that @xmath178 the estimate for @xmath72 follows from the same strategy as in the real case .
first of all , just like the real case for any @xmath96 unitary matrix @xmath179 the entries of @xmath180 are still independed complex standard normal random variables . as a result the probability @xmath181 ) where @xmath182 is a unit vector does not depend on the choice of @xmath21 . by taking @xmath99
we see that @xmath183 has the @xmath184 distribution ( as opposed to the @xmath185 distribution in the real case ) .
applying lemma [ lem-2.1 ] we obtain the equivalent result for the complex case in @xmath186 next for the @xmath111-net , we observe that the unit sphere in @xmath50 is precisely the unit sphere in @xmath187 if we identify @xmath50 as @xmath187 .
thus we can find an @xmath111-net @xmath113 of cardinality no more than @xmath188 .
the proof of theorem [ theo-1.2 ] now goes through with some minor modifications .
the most important one is that with ( [ 2.16 ] ) and ( [ 2.17 ] ) the inequality condition ( [ 2.7 ] ) now becomes @xmath189 where the constant @xmath137 is changed to @xmath177 . substituting this @xmath137 and @xmath190 for @xmath80 we prove the theorem . ' '' ''
our goal in this section is to establish the robustness of random frames against erasures by proving theorem [ theo-1.1 ] .
here we restate theorem [ theo-1.1 ] in a a different form for the benefit of simpler notation in the proof .
[ theo-3.1 ] let @xmath44 where @xmath45 is @xmath13 whose entries are drawn independently from the standard normal @xmath46 distribution .
let @xmath61 and @xmath191 where @xmath192 .
for any @xmath63 there exist constants @xmath64 depending only on @xmath81 , @xmath193 and @xmath66 such that @xmath14 is a @xmath194-nerf with probability at least @xmath195 .
* proof .
* there exists exactly @xmath196 subsets @xmath197 of cardinality @xmath198 .
it is well known that @xmath199 which can be shown easily by stirling s formula or induction on @xmath49 .
set @xmath200 , which has @xmath201 .
we have then @xmath202 now we set @xmath203 .
let @xmath204 and @xmath84 where @xmath156 is given in ( [ 1.4 ] ) .
let the columns of @xmath45 be @xmath205 .
for any @xmath197 we denote by @xmath206 the submatrix of @xmath45 whose columns are @xmath207 .
then for @xmath208 we have @xmath209 by theorem [ theo-1.2 ] .
it follows that @xmath210 it follows that @xmath211 this implies that , by setting @xmath212 and @xmath213 , @xmath44 is a @xmath36-nerf with probability at least @xmath195
. theorems [ 1.1 ] and [ theo-3.1 ] states that random gaussian frames can be robust with overwhelming probability against erasures of an arbitrary proportion of data from the original data , at least in theory , as long as the number of remaining vectors is at least @xmath214 for some @xmath1 . in practice one may ask how good the condition numbers are if the erasures reach a high proportion , say , 90% of the data .
we show some numerical results below
. * example 1 .
* let @xmath215 where @xmath45 is @xmath13 whose entries are independent standard normal random variables .
set @xmath216 . in this experiment
we fix @xmath217 and @xmath218 , respectively , and let @xmath49 vary .
as @xmath49 increases from @xmath219 to @xmath220 the proportion of erasure @xmath221 increases from 0 to 99% .
we shall use @xmath222 as a measure of robustness since it is an upper bound for the condition number . clearly , as @xmath223 increases we should expect @xmath222 to increase . the left plot in figure [ fig-1 ]
shows @xmath224 against @xmath223 for both @xmath217 ( top curve ) and @xmath218 ( bottom curve ) .
because the frame is normalized so that each column is on average a unit norm vector , it also makes sense to use the smallest singular value as a measurement of robustness .
the right plot in figure [ fig-1 ] shows @xmath225 against @xmath223 also for both @xmath217 ( top curve ) and @xmath218 ( bottom curve ) .
our numerical results show that in the case @xmath217 , with probability at least @xmath226 , the condition number is no more than @xmath227 for @xmath52% erasures and no more than @xmath228 for 90% erasures . in the case
@xmath218 , the corresponding numbers are 139.88 and 1862.1 , respectively .
in fact , even with 99% erasures the condition number is no more than 42716 . left : @xmath224 against the proportion of erasures when @xmath49 varies from @xmath47 to @xmath229 while @xmath47 is fixed at @xmath217 ( top curve ) and @xmath218 ( bottom curve ) .
right : same as in the left figure , but for @xmath225.,title="fig:",width=272,height=226 ] left : @xmath224 against the proportion of erasures when @xmath49 varies from @xmath47 to @xmath229 while @xmath47 is fixed at @xmath217 ( top curve ) and @xmath218 ( bottom curve ) .
right : same as in the left figure , but for @xmath225.,title="fig:",width=272,height=226 ] * example 2 .
* again we let @xmath230 where @xmath45 is @xmath13 whose entries are independent standard normal random variables , and let @xmath231 .
in this experiment we fix @xmath232 and @xmath233 , respectively , and let @xmath47 vary so the proportion of erasures @xmath234 varies from 0 to 99% ( @xmath232 and 0 to 97% ( @xmath233 ) , respectively .
again we should expect the robustness to go down as we increase @xmath223 . the left plot in figure [ fig-2 ] shows @xmath224 against @xmath223 for @xmath233 ( top curve ) and @xmath232 ( bottom curve ) .
the right plot in figure [ fig-2 ] shows @xmath225 against @xmath223 also for both @xmath233 ( top curve ) and @xmath232 ( bottom curve ) .
our numerical results show that in the case @xmath233 , with probability at least @xmath226 , the condition number is no more than @xmath235 for @xmath52% erasures and @xmath236 for 90% erasures .
in the case @xmath232 , the corresponding numbers are 23.48 and 315.12 , respectively .
even with 95% erasures the condition number is no more than 1312.4 .
left : @xmath224 against the proportion of erasures when @xmath47 varies while @xmath49 is fixed at @xmath233 ( top curve ) and @xmath232 ( bottom curve ) .
right : same as in the left figure , but for @xmath225.,title="fig:",width=272,height=226 ] left : @xmath224 against the proportion of erasures when @xmath47 varies while @xmath49 is fixed at @xmath233 ( top curve ) and @xmath232 ( bottom curve ) .
right : same as in the left figure , but for @xmath225.,title="fig:",width=272,height=226 ] | data erasure can often occur in communication . guarding against erasures involves redundancy in data representation .
mathematically this may be achieved by redundancy through the use of frames .
one way to measure the robustness of a frame against erasures is to examine the worst case condition number of the frame with a certain number of vectors erased from the frame . the term _ numerically erasure - robust frames ( nerfs ) _ was introduced in @xcite to give a more precise characterization of erasure robustness of frames . in the paper
the authors established that random frames whose entries are drawn independently from the standard normal distribution can be robust against up to approximately 15% erasures , and asked whether there exist frames that are robust against erasures of more than 50% . in this paper
we show that with very high probability random frames are , independent of the dimension , robust against any amount of erasures as long as the number of remaining vectors is at least @xmath0 times the dimension for some @xmath1 .
this is the best possible result , and it also implies that the proportion of erasures can arbitrarily close to 1 while still maintaining robustness .
our result depends crucially on a new estimate for the smallest singular value of a rectangular random matrix with independent standard normal entries . |
binary systems have long been used as unique probes into the origins of stars , as the dynamical imprint of their formation and subsequent evolution are recorded in their orbits .
in fact , decades before it was known that nucleosynthesis powers stars , well - determined binary orbits were commonplace ( e.g. , @xmath7 orbits were published in the compilation of * ? ? ?
thus , the study of binary orbits has a rich heritage , predating much of modern astrophysics .
however , only very recently has it been possible to measure orbits for the lowest mass products of star formation : stars at the bottom of the main - sequence ( @xmath8 ) and brown dwarfs .
the sample of orbits for such objects has been growing rapidly in recent years as astrometric monitoring has begun to yield orbits for the shortest period visual binaries known ( @xmath9 years ) , which were discovered by high - resolution imaging surveys 510 years ago ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
some properties of very low - mass binaries , such as orbital semimajor axes , mass ratios , and multiplicity fractions , can be studied in a statistical fashion using a large sample without any knowledge of orbital motion ( e.g. , * ? ? ?
* ; * ? ? ?
however , only well - determined orbits can provide a measure of eccentricity ( @xmath5 ) , which is directly related to the dynamics of the formation process . because of the long history of stellar orbit studies , many authors have weighed in on the subject of the eccentricity distribution .
the most recent comprehensive work on the eccentricities of solar - type field binaries was done by , and their pre main - sequence counterparts have been reviewed by .
now for the first time it is possible to bring similar observations to bear on the extreme low - mass end of the binary population .
some of the proposed formation mechanisms for such low - mass objects , such as ejection ( e.g. , * ? ? ?
* ) and gravitational instability ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , may leave a distinct signature in the orbits of binaries . here
we present the first comprehensive look at what orbital eccentricity measurements reveal about the nature of very low - mass stars and brown dwarfs .
we first updated published orbit determinations for very low - mass binaries that have new data before compiling our sample for this study . for two visual binaries ( hd 130948bc and lp 415 - 20ab )
, we present new astrometry from our ongoing keck adaptive optics ( ao ) monitoring program in order to refine their orbital elements .
our procedure for obtaining , reducing , and analyzing our natural guide star and laser guide star ao images is described in detail in our previous work ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , and the binary parameters derived from our new observations are shown in table [ tbl : astrom ] .
we used the markov chain monte carlo ( mcmc ) approach described by @xcite to determine the posterior probability distributions of the orbital elements for these binaries , and in table [ tbl : orbits ] we give their median and 68.3% confidence limits . for hd 130948bc
, we combined our new astrometry with the epochs published by @xcite .
we updated this astrometry to account for the newer astrometric calibration of nirc2 by @xcite , which amounts to a @xmath10 offset to our previously published position angles . the extended orbital phase coverage results in a factor of @xmath114 improvement in precision compared to the published orbital parameters .
some binaries have multiple orbit determinations in the literature , and in these cases we used our published results in the following analysis . in most cases published
orbits are in good agreement with each other , but a few binaries warrant special mention . in the case of lp 415 - 20ab , our astrometry results in a very different orbit than found by @xcite . as a test
, we fitted their three keck epochs with the published gemini epoch from @xcite and found an orbit consistent with the one they reported ( @xmath12 ) .
however , after including our four keck epochs we found that the 2007 december astrometry of @xcite is highly discrepant with the ensemble of measurements , resulting in a @xmath13 of 21.5 for 9 degrees of freedom ( dof ) .
if we exclude this one measurement , we find good agreement between the data sets , and the best - fit orbit has a reasonable @xmath13 ( 6.83 for 7 dof ) and significantly different orbital parameters ( e.g. , our newly derived eccentricity of @xmath14 is 1.9@xmath15 discrepant with their @xmath16 ) .
it appears that with only four epochs ( i.e. , 1 dof ) the orbit fit of @xcite was vulnerable to a single discrepant astrometric measurement leading to an incorrect orbit solution .
we also note that our published orbit for 2mass j2206@xmath172047ab @xcite is nominally inconsistent with that published by @xcite , with all orbital parameters discrepant by @xmath1124@xmath15 . using only their data ,
we have found a lower @xmath13 orbit solution than they report , and although it is more consistent with our published orbit , it still has an unrealistically large @xmath13 ( 30.3 for 9 dof ) .
thus , we choose to use our published orbit for 2mass j2206@xmath172047ab ( @xmath18 for 7 dof ) in the following analysis .
finally , the orbit for 2mass j1534@xmath172962ab published by @xcite is marginally inconsistent with the original orbit from @xcite . with the latest astrometry
the period and semimajor axis are smaller by 1.8@xmath15 , while the eccentricity and mass are consistent within 1@xmath15 .
we use the more recent orbit from @xcite in the following analysis .
we have constructed a sample of very low - mass ( @xmath19 ) binary orbits , including both visual and spectroscopic binaries . because it is imperative for our purposes that this sample be as unbiased as possible with respect to eccentricity ,
in the following analysis we consider potential sources of selection bias , the nature of which varies between different types of binaries .
below we describe how we have selected our orbit sample as well as the tests we have performed to quantify the level of eccentricity bias present in this sample . in table
[ tbl : vb ] , we present a list of all 20 visual binaries that have published orbit determinations and a total system mass @xmath20 . for each binary
we list its distance , semimajor axis ( @xmath21 ) , and eccentricity ( @xmath5 ) , as these are the physical parameters that , along with three viewing angle parameters , determine the size and shape of a binary s orbit on the sky .
thus , these are the three key parameters that must be considered when characterizing selection biases . in most cases
we have used orbit parameters as published and distances determined from trigonometric parallax measurements . when parallax measurements were not available , we estimated distances based on the primary components resolved photometry and spectral types using the updated spectral type
absolute magnitude relations described in appendix b of @xcite . for all orbits published by other authors ,
we have created our own markov chains using the published astrometry and the same analysis method described in section [ sec : orbits ] . however , only in the cases where the astrometric orbit fits yield a reasonable @xmath13 can we trust the mcmc analysis to provide accurate posterior distributions of the orbital parameters . in cases where @xmath13 is unreasonably large ,
the parameter space is not fully explored and uncertainties can be significantly underestimated by the mcmc method ( or in fact by any fitting method ) .
we find such large @xmath13 values ( reduced @xmath22120 ) for five of the orbits published by @xcite . for three of these orbits they combined astrometry with radial velocities ( 2mass j1847@xmath235522ab , 2mass j1750@xmath234424ab , 2mass j1426@xmath231557ab ) , which may have enabled them to achieve orbit fits with smaller @xmath13 than we found using the astrometry alone .
for the other two orbits ( 2mass j0850@xmath231057ab and 2mass j1728@xmath233948ab ) , it is unclear why we are unable to reproduce their lower @xmath13 values from the same input data ( astrometry alone ) .
for all five of these orbits we use the parameters derived by @xcite .
( note that for reasons unrelated these discrepancies , none of these five binaries ultimately passes the selection criteria that we define below for our unbiased eccentricity sample . ) with a complete set of the known very low - mass visual binary orbits in hand , we now consider how selection biases may affect the eccentricity distribution of this sample .
visual binaries must pass three independent selection criteria before having their orbits determined : ( 1 ) they must be initially resolved by a high - resolution imaging survey ;
( 2 ) they must be included in subsequent astrometric monitoring programs ; and ( 3 ) their observed relative motion must yield a unique orbit fit .
the first selection cut can result in biases on orbital parameters , as has been shown for direct imaging searches for extrasolar planets ( e.g. , see @xcite ) .
the second cut results from the fact that it is impractical to monitor the orbits of all @xmath11100 known very low - mass binaries .
rather , astrometric monitoring programs are inevitably optimized for measuring as many orbits as possible with limited observing resources over a limited amount of time by focusing on the binaries likely to have the shortest orbital periods .
thus , the selection criteria used to define the monitoring sample can result in strong biases on the orbit parameters . because the orbital period must be estimated from the projected separation of the binary , binaries with wide separations are much less desirable to monitor .
this introduces an eccentricity bias because more eccentric binaries at a given period and mass will , on average , have wider projected separations .
we will therefore attempt to define a subset of visual binaries that has minimal eccentricity bias due to these selection criteria .
* @xmath24 discovery bias .
* we first consider the effect of `` discovery bias '' which causes the population of binaries resolved by imaging surveys to deviate from the true underlying eccentricity distribution of all binaries .
for a given imaging surveys , the extent of this bias depends on how the semimajor axis ( @xmath21 ) distribution compares to that survey s resolution limit , or inner working angle ( iwa ) .
for example , if the @xmath21 distribution peaks inside the iwa , the discovery of higher eccentricity binaries is favored , as such binaries spend more time at wider separations compared to more circular orbits . to quantify this bias , we have performed a simple monte carlo simulation in which @xmath21 was held fixed , while eccentricity was distributed uniformly from 0 to 1 , viewing angles were drawn from appropriate random distributions ( @xmath25 ; uniform in @xmath26 and @xmath27 ) , and the single - epoch observations were distributed uniformly in orbital phase .
binaries were considered detected if their projected separation at the time of observation was above the resolution limit .
the impact on the uniform input eccentricity distribution is illustrated in figure [ fig : bias ] .
when the resolution limit falls at or outside the fixed semimajor axis ( iwa @xmath28 ) , high-@xmath5 orbits are strongly preferred as they spend the most time wider than the resolution limit .
when the resolution limit falls inside the fixed value of @xmath21 ( iwa @xmath29 ) both highly eccentric and circular orbits are preferred , with intermediate eccentricities being selected against .
this is because in these cases any sufficiently eccentric orbit can be made unresolvable by an unfavorable viewing angle , but this can be compensated by the fact that the most eccentric orbits spend extended periods of time at very wide separations .
thus , high-@xmath5 orbits are more likely to be detected than intermediate-@xmath5 orbits , as they spend the most time widely separated .
to predict the exact form of the eccentricity discovery bias for our sample would require an assumption for the underlying semimajor axis distribution as well as careful accounting of the resolution limits of the numerous imaging surveys that have contributed to the current sample of very low - mass binaries .
however , for our purposes we simply want to ensure that our sample of orbits has an acceptably small amount of bias .
thus , we need only chose a cutoff in the ratio of iwa/@xmath21 that is small enough to minimize the bias while including as many binaries with orbit determinations as possible . in table
[ tbl : vb ] , we give our estimate for the iwa of the imaging data used to discover each visual binary as well as the ratio iwa/@xmath21 . we have chosen a cutoff of iwa/@xmath30 , which figure [ fig : bias ] shows produces a relatively unbiased selection over a full range of eccentricities from 0 to 1 .
this cutoff excludes three binaries : lp 415 - 20ab ( iwa/@xmath31 ) , 2mass j0920@xmath233517ab ( iwa/@xmath32 ) , and 2mass j2140@xmath231625ab ( iwa/@xmath33 ) .
the highest remaining iwa/@xmath21 value in our sample is 0.68 for lp 349 - 25ab .
while it is unfortunate to have to exclude any orbits from our sample , we emphasize that this is a very important step in keeping potentially large biases from impacting the following analysis .
the case of lp 415 - 20ab is a prime example of the problem with discovery bias on eccentricity .
this binary was discovered at a projected separation of @xmath34 mas , right at the resolution limit of the imaging survey of @xcite .
its actual semimajor axis turns out to be @xmath35 mas , so it could have only been discovered by that survey if it had an eccentric orbit , as it indeed does ( @xmath36 ) .
thus , while lp 415 - 20ab does tell us that very low - mass binaries can be quite eccentric , it can not convey any other information about the eccentricity distribution since the only reason it was discovered is because of its large eccentricity .
* @xmath24 monitoring sample bias .
* next , we consider how to prevent the sample selection of orbital monitoring programs from biasing the resulting eccentricity distribution .
this is essentially a matter of choosing a range of semimajor axes for which the monitoring sample is complete at all eccentricities .
the lower limit in @xmath21 is simply defined by the visual binary orbit with the smallest semimajor axis ; this is gl 569bab with @xmath37 au @xcite .
there are also no known visual binaries with projected separations ( @xmath38 ) at discovery smaller than 0.9 au , so we define this as the lower limit of our sample .
the upper limit should be chosen to accurately reflect the completeness of orbit monitoring surveys that tend to focus on the smallest separation binaries , while also including as many of the resulting orbit measurements as possible .
for the many binaries without orbit determinations , we must use their minimum possible semimajor axis , which would be the case of a highly eccentric , face - on orbit caught at apoastron ( physical separation of @xmath39 ) . since bound orbits have @xmath40 , the minimum possible semimajor axis for any binary is therefore half its projected separation ( @xmath41 ) .
thus , some binaries that seem to be very wide ( i.e. , poor monitoring targets ) could in fact have much smaller semimajor axes but just be very eccentric .
when choosing an upper limit in @xmath21 , we must therefore ensure that orbit monitoring programs are complete up to projected separations of @xmath42 , otherwise we would introduce a strong selection bias against high eccentricities .
we choose an upper limit in @xmath21 of 3.7 au in order to include the binaries lhs 1070bc ( @xmath43 au ; ) and lhs 1901ab ( @xmath44 au ; @xcite ) .
the next widest binary that would also pass our iwa/@xmath21 cut is 2mass 0850@xmath231057ab ( @xmath45 au , @xmath46 ) . our chosen 3.7 au upper limit means that binaries up to projected separations of 7.4 au should not have been intentionally excluded from orbit monitoring programs . to the best of our knowledge
this is the case , as at least a second epoch has been obtained for all binaries with @xmath47 au , either by our own keck program or others .
a slightly larger upper limit of 4.0 au would require several more binaries to have undergone orbit monitoring that have not , while adding no additional binaries with orbit determinations to our sample .
* @xmath24 monte carlo simulations . * with these selection criteria of @xmath48 au , @xmath47 au , and iwa@xmath49 in hand , we have performed monte carlo simulations to assess the level of eccentricity bias that is expected for our sample . for this range of semimajor axes , the very low - mass binary parameter distributions derived by @xcite are appropriate .
the semimajor axis distribution is lognormal , with a peak at @xmath50 and @xmath51 . to convert @xmath21 from au to arcseconds
, we randomly drew distances in a monte carlo fashion incorporating the selection effects of the seeing - limited surveys that originally discovered the binaries as unresolved sources .
we began with objects distributed uniformly in volume .
we assigned magnitudes to these objects according to the empirical -band luminosity function of @xcite which is valid from @xmath52 mag ( i.e. , @xmath11m8 to l8 , encompassing the spectral type range of the bulk of our sample ) .
we further added flux due to unresolved companions according to the @xcite binary frequency of 20% and mass ratio distribution of @xmath53 ; we converted mass ratio to flux ratio using the @xmath54 theoretical power - law relation of @xcite .
the final distance distribution was then determined by magnitude cuts on this simulated population .
we tested both strong ( @xmath55 mag ) and weak ( @xmath56 mag ) magnitude cuts equivalent to those used by @xcite and @xcite in mining 2mass for late - m dwarfs and l dwarfs , respectively . in the latter case , we applied an additional @xmath57 mag cut to approximate the @xmath58 mag color cut of @xcite that excluded brighter , bluer m dwarfs ; we based this cut on the color
absolute magnitude diagrams of @xcite .
the strong magnitude cut for the brighter late - m sample resulted in a distance distribution peaking at @xmath1120 pc , while the weaker cut for the fainter l dwarf sample resulted in similarly shaped distribution peaked at @xmath1140 pc .
we then randomly generated an observed projected separation ( @xmath38 ) for each binary in a similar manner as described earlier ( i.e. , random viewing angles and observing times , and a uniform eccentricity distribution ) .
figure [ fig : bias ] shows the resulting eccentricity distribution after applying the selection criteria of @xmath48 au , iwa@xmath49 , @xmath59 au , and @xmath60 ( i.e. , the binary was resolved ) . for both samples we conservatively assumed the worst - case iwa from the surveys that originally discovered the binaries in our sample . for
the late - m dwarfs this is 0@xmath6112 from the gemini ao surveys ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , and for the l dwarfs this is 0@xmath6107 from both _ hst _ and keck surveys ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . by assuming the worst - case iwa we find the largest possible amplitude of the eccentricity bias , since smaller values of the iwa will always yield less bias ( i.e.
, more binaries will always be uncovered for a smaller iwa ) .
the resulting maximal bias we find from our monte carlo simulations has an amplitude of less than @xmath62 for both the late - m and l samples .
as expected , the level of eccentricity bias introduced by our selection criteria is negligible , since they were chosen to minimize the eccentricity bias , and as shown in figure [ fig : bias ] there is no strong systematic trend in the bias .
nearly circular ( @xmath63 ) and highly eccentric orbits ( @xmath64 ) are slightly preferred ( by @xmath114% ) as compared to intermediate eccentricities .
we neglect this bias in the following analysis
. * @xmath24 orbit fitting bias . *
finally , we consider one additional selection effect present in our sample : each binary must have undergone astrometric motion suitable for yielding a robust orbit determination .
this can potentially introduce an eccentricity bias if orbit fits are more easily determined for certain ranges of eccentricity .
for example , it is easy to imagine that an extremely eccentric binary would present a challenge to astrometric monitoring as it spends most of its time at a fixed projected separation before moving rapidly inward in nearly a straight line
. such an orbit would be difficult to observe at just the right times needed for a good fit and would also appear to be less appealing for continued monitoring .
@xcite investigated this bias by generating @xmath65 random orbits and classifying the expected quality of the orbit fit by eye , relying on their extensive experience working with such astrometric data .
such an ad hoc method is necessary because in the marginal cases of visual binary orbit fitting the process of @xmath13 minimization is highly nonlinear and strongly dependent on the initial guess .
@xcite found that on average orbits with @xmath66 all had the same likelihood of yielding a good fit , so they assigned this eccentricity range a normalized completeness factor of 100% .
they found that this completeness drops to 77% for eccentricities of 0.60.8 and 42% for @xmath67 . in our sample
, we have one visual binary with @xmath64 and none in the 0.60.8 range . in the following analysis
we consider the impact of applying correction factors of 1.3@xmath68 and 2.4@xmath68 due to this orbit fitting bias . *
@xmath24 final visual binary sample . * in table [ tbl : ecc ] , we list the orbital periods , eccentricities , and masses of the eleven visual binary orbits from table [ tbl : vb ] that meet our eccentricity bias - minimizing criteria of having @xmath48 au and a discovery iwa@xmath49 .
( note that we list orbital period here rather than semimajor axis in order to compare these orbits directly to the unresolved spectroscopic binary orbits . )
although this nominally includes only a little over half of the 20 published visual binary orbit determinations , in fact the majority of the remaining orbits would not contribute much information as they are not as well constrained ( see the orbit quality metrics given in table [ tbl : vb ] ) .
for example , all orbits in our bias - minimized sample happen to have 1@xmath15 eccentricity confidence limits of @xmath69 ( median 0.005 ) . however , of the nine orbits not included in the sample , six of them have much more poorly constrained orbits ( 1@xmath15 confidence limits of 0.10.5 ) , with a median error of 0.26 that is @xmath70 larger than the median for the other orbits .
in fact , the reason why these six orbits are poorly determined is the same as why they were excluded by our bias minimizing criteria : they are all the widest ( i.e. , longest period ) binaries , and this results in them having orbit monitoring observations that cover much less orbital phase ( as little as 0.2% and 1.3% of the orbit is covered in the two most extreme cases ) . to date ,
five unresolved very low - mass binaries have published orbit determinations .
two of these were discovered via radial velocity ( rv ) monitoring as single - lined spectroscopic binaries ( sb1s ) : cha h@xmath71 8 ( @xmath72 ; * ? ? ?
* ) , and 2mass j0320@xmath170446 ( @xmath73 ; * ? ? ?
@xcite have examined the selection bias introduced by rv surveys on the eccentricity distribution through monte carlo simulations .
they found that as long as the rv semiamplitude ( @xmath74 ) of an sb1 is @xmath753@xmath68 larger than the rms of the residuals ( @xmath15 ) , its eccentricity does not greatly impact its likelihood of discovery .
as shown in their figure 4 , sb1s have a constant detection efficiency up to @xmath76 , beyond which there is a slight downturn .
however , even at the highest eccentricities ( @xmath77 ) the detection efficiency is only @xmath1120% less than for @xmath78 orbits .
the reason for this lack of a strong eccentricity bias in rv data is because eccentricity acts to boost the rv semiamplitude ( @xmath79 ) , counteracting the fact that eccentric orbits are more difficult to adequately sample with times series data .
the two sb1s in our sample , cha h@xmath71 8 and 2mass j0320@xmath170446 , have large semiamplitudes relative to their rms ( @xmath80 and 21 , respectively ) , and thus are expected to harbor little bias ( @xmath81 ) with respect to eccentricity according to the analysis of @xcite . for this reason and the small size of the sample ( 2 systems ) we neglect the eccentricity bias for sb1s .
one double - lined spectroscopic binary ( sb2 ) has a measured orbit : ppl 15 ( @xmath82 * ? ? ?
it was discovered as an sb2 in three epochs of rv data .
the identification of sb2s in spectroscopic data is subject to different eccentricity selection effects than for sb1s because changes in the velocity do not need to be measured to detect the two sets of absorption lines due to the binary . to quantify the eccentricity bias in detecting sb2s , we performed a monte carlo simulation similar to the one described above for visual binaries but with the detection criterion that the velocity shift between the binary components must be large enough to detect in high - resolution spectra at one of three random epochs .
we fixed the period and semimajor axis to that of ppl 15 ( 5.8 days , 0.03 au ) .
does not depend on this quantity ( @xmath83 ) .
of course , very low mass ratio systems of all eccentricities are equally selected against as the secondary s absorption lines are more difficult to detect in the integrated - light spectrum . ]
figure [ fig : bias - sb2 ] shows the resulting sb2 eccentricity bias predicted by our simulation for a detection threshold of @xmath84 ( i.e. , 3@xmath68 the spectral resolution of @xcite ) .
we find only a slight ( @xmath113% ) preference for the discovery of highly eccentric ( @xmath67 ) sb2s in such rv data . for comparison
, we also show the results if the detection threshold were higher or lower by a factor of two , as would be the case for binaries with rv semiamplitudes different by a factor of two .
higher semiamplitude binaries are less biased , while for lower semiamplitudes the eccentricity bias becomes a larger effect ( though still @xmath85 ) .
thus , it seems reasonable to neglect eccentricity bias for ppl 15 , the one sb2 in our orbit sample .
lsr j1610@xmath170040 ( @xmath86 ) is the only very low - mass astrometric binary known to date , discovered by the usno parallax program @xcite .
it was later found to be an sb1 in the multi - epoch rv data obtained by @xcite . in principle ,
astrometric time series data is very similar to rv time series data and so would be subject to the same biases as for sb1s .
thus , we assume that the same selection effects apply , namely that orbits with @xmath87 have an @xmath1180% detection efficiency relative to less eccentric orbits , which are all equally accessible by astrometric monitoring . finally
, just one very low - mass eclipsing binary is known , 2mass j0535@xmath170546 ( @xmath88 ) , discovered by @xcite in a photometric variability survey . in a study of exoplanet transit surveys
, @xcite showed that photometric transit / eclipse detections are only slightly biased ( @xmath89 ) in eccentricity , with the exact correction depending on the underlying @xmath5 distribution .
thus , we neglect any eccentricity bias for 2mass j0535@xmath170546 . table [ tbl : ecc ] lists the orbital periods , eccentricities , and masses for these five unresolved binaries orbits . combined with the visual binary sample ,
our sample comprises a total of 16 very low - mass binary orbits spanning orbital periods from 5.8 days to 8400 days .
the eccentricity distribution and orbital period eccentricity diagram for our sample is shown in figure [ fig : vlm ] .
our sample shows a strong preference for modest eccentricities , with a median eccentricity of 0.34 ( mean also 0.34 ) for the full sample and a median of 0.33 ( mean of 0.33 ) for the visual binaries alone .
the visual binary orbits show a broad range of eccentricities , from near - circular up to @xmath90 . at the shortest periods ( @xmath91 days ) ,
we find an absence of eccentric orbits ( @xmath92 ) , although this could be due to small number statistics .
thus , we find no strong evidence that the eccentricity distribution of very low - mass binaries changes over three orders of magnitude in orbital period ( @xmath93 days ) . we have quantified the lack of a correlation between period and eccentricity by computing spearman s rank correlation coefficient ( @xmath94 ) for our sample .
we use a monte carlo approach , accounting for the period and eccentricity errors by drawing the data in each trial from the observed distributions ( our mcmc chains when possible , gaussians otherwise ) . the resulting mean and rms of the correlation coefficient
was @xmath95 , which is well below the critical value of 0.412 needed to show correlation for our sample even at 90% confidence .
the subset of visual binaries also shows no significant @xmath96@xmath5 correlation ( @xmath97 , critical value 0.497 at 90% confidence ) or @xmath21@xmath5 correlation ( @xmath98 ) .
in contrast , @xcite reported a trend in @xmath5 as a function of period for their sample of fifteen very low - mass visual binaries , although they did not quantify the significance of the correlation .
we note that their observed trend is largely dependent on the two longest period orbits in their sample ( @xmath99 yr and @xmath100 yr ) for which they found large eccentricities with large errors ( @xmath101 and @xmath102 , respectively ) .
both of these binaries were excluded from our sample by our criteria for minimizing eccentricity bias , and we also note that with only 2.2% and 0.4% observational phase coverage these orbit determinations and their uncertainties are not likely to be robust ( e.g. , see the cautionary example of df tau ab in @xcite ) .
we have also investigated whether mass ratio is correlated with eccentricity for our sample . in table
[ tbl : ecc ] , we list the best available mass ratio ( @xmath103 ) estimates or measurements for our sample binaries . for the visual binaries , @xmath104 is estimated from lyon evolutionary models given the measured total mass and individual luminosities using the method described in section 5.4 of @xcite .
this has been done in a consistent fashion for the sample presented here , so even if there are systematic errors in the evolutionary models the relative ordering of the mass ratios should be relatively unaffected .
ranges typically spanning a factor of @xmath105 in mass ratio except for gl 569bab ( @xmath106 , i.e. , 0.580.90 at 1@xmath15 ) , which is still very large for our purposes .
thus , we have used our model - based @xmath104 estimates for these binaries . ] for the eclipsing binary 2mass j0535@xmath170546 , we used its directly measured mass ratio ( @xmath107 ; * ? ? ?
none of the three sb1s in our sample has a well - determined mass ratio , so we excluded them from our mass ratio analysis .
we also excluded one visual binary ( lhs 2397aab ) because it has a very large uncertainty in its mass ratio estimate ( @xmath108 ; * ? ? ? * ) due to the fact that it is composed of a main - sequence star and a field brown dwarf .
we computed spearman s rank correlation coefficient in a monte carlo fashion as described above and found @xmath109 , showing that there is no evidence for a correlation of eccentricity with mass ratio over the range of our sample ( @xmath110 ) . in the absence of such a correlation ,
any mass ratio selection effects present in our sample should therefore not induce a systematic eccentricity bias .
have examined in detail the eccentricity distribution of a well - defined sample of solar - type binaries . in the following analysis
, we exclude from consideration three of the binaries in their sample that have an orbit quality grade of 4 or 5 assigned by @xcite , which indicates an indeterminate orbit . not surprisingly , these binaries ( hd 16895ab , hd 78154ab , and hd 146361ab ) are the longest period orbits in the sample , with orbital periods @xmath111 yr .
( we note that @xcite have recently updated this volume - limited , solar - type sample , but we can not use this for comparison since they did not publish the eccentricity values of their sample binaries . )
we find that the overall eccentricity distribution of the solar - type binary sample is very similar to our orbit sample . to assess this quantitatively , we performed kolmogorov - smirnov ( ks ) tests in a monte carlo fashion .
we accounted for eccentricity errors in our sample by drawing the data in each trial from the observed distributions ( our mcmc chains when possible , gaussians otherwise ) .
we computed the ks statistic ( @xmath112 ) and the corresponding probability for every trial .
the resulting rms scatter in @xmath112 was relatively small(@xmath113 ) .
is the maximum difference between the two cumulative distribution functions that are being tested . ] for the final probability that two samples were drawn from the same parent distribution , we used the mean of the probabilities from individual trials .
table [ tbl : ks ] shows the resulting probabilities comparing subsamples of our orbits to subsamples of the solar - type binaries from .
in general , we find good agreement between the two samples eccentricity distributions .
however , a more detailed examination of these samples shows some significant differences . unlike the very low - mass binaries ,
solar - type binaries follow a strong correlation with orbital period such that short - period ( @xmath114 days ) binaries span a different range of eccentricities than do long - period ( @xmath115 days ) binaries ( figure [ fig : compare ] ) . suggested that this implies a different formation mechanism and/or dynamical history for these two period regimes of solar - type binaries .
the same trend has also been reported for pre
main - sequence , solar - type binaries , implying that this period eccentricity correlation is determined early in a binary s formation history .
we confirm the significance of the correlation in the sample by computing spearman s rank correlation coefficient for solar - type binaries ( excluding those that lie below the critical period for tidal circularization ; @xmath116 days ) .
the resulting value of @xmath117 for this sample of 46 solar - type binaries gives a 98.6% confidence that @xmath96 and @xmath5 are correlated .
as discussed above , our sample shows no such correlation , which is an indication that the solar and very low - mass populations are in fact dynamically distinct from one another .
we also note that among the large sample of solar - type binaries from there is not a single orbit with @xmath3 beyond an orbital period of 22 days , days and @xmath3 are indeed known to exist , but none are included in the well - defined , volume - limited sample of .
for example , in the sb9 catalog 20% of orbits with @xmath118 days and quality grades @xmath119 have @xmath3 .
however , such spectroscopic binary catalogs represent a compilation of heterogeneous data not intended to be used for statistical studies , as they likely harbor significant selection biases ( e.g. , as discussed in section [ sec : sb ] ) .
in addition , we note that the @xcite solar - type sample contains 3 such binaries according to their figure 14 .
] while there are at least 3 such orbits in our sample of 16 binaries ( one more , 2mass j1534@xmath172952ab , could also have @xmath3 , but the error in its eccentricity makes this uncertain ) . this feature of solar - type systems has also been pointed out for pre main - sequence binaries .
we used a fisher s exact test to compute the probability that this difference in @xmath3 orbits between our sample and the sample arises solely by chance ( i.e. , the @xmath96-value ) . in order to exclude any solar - type binaries that may be affected by tidal circularization but have not reached @xmath120 yet
, we consider only binaries with @xmath118 days ( a less conservative , shorter period cutoff could be used but would not change the results ) . at these orbital periods , 0 of 39 solar - type binaries have @xmath3 , while 3 of 14 very low - mass binaries have eccentricities confidently @xmath8 .
comparing these two samples , fisher s exact test finds that this difference is significant with a @xmath96-value of 0.0155 ( 2.2@xmath15 ) .
the @xmath96-value remains significant ( @xmath121 ) for any choice of the eccentricity division point ranging from 0.10 to 0.15 .
this strong preference for near - circular orbits among very low - mass binaries provides additional evidence that their dynamical history may be different from solar - type binaries .
we next consider the location in the @xmath96@xmath5 diagram of very low - mass binaries that are members of hierarchical triple systems as compared to their solar - type counterparts .
the interpretation given by and @xcite is that solar - type binaries that are members of hierarchical triple or quadruple systems tend to lie near the upper envelope of eccentricity at a given orbital period .
however , examination of the @xmath96@xmath5 diagram of solar - type binaries presented in figure 14 of @xcite shows that such members of hierarchical systems actually occupy a wide range of eccentricities at periods @xmath122 days .
in fact , members of hierarchical triple and quadruple systems are the only near - circular ( @xmath123 ) solar - type binaries at these periods , suggesting that their higher order nature may be the cause of their unusually low eccentricities .
this is consistent with the fact that the only such long - period , near - circular binary in the pre main - sequence sample of is now known to have a tertiary companion ( gw ori ; * ? ? ?
in contrast to solar - type binaries , the higher order multiples in our sample do not populate the upper eccentricity envelope at any period .
in fact , the highest eccentricity orbit in our sample is a true binary , and the lowest eccentricity orbit belongs to a member of a triple system .
this latter datum may be partially consistent with the fact that near - circular ( @xmath3 ) solar - type binaries tend to be members of higher order multiples .
however , there remains the mystery as to why such near - circular orbits are much more common among the true binaries in our sample than in the solar - type binary sample .
finally , we note that the shortest period binaries in our sample are not circularized , despite having shorter periods than the empirical limit from field solar - type binaries ( @xmath124 days ) .
this was discussed by @xcite , who used brown dwarf structural models along with the circularization theory of to show that ppl 15 ( @xmath125 days ) is not expected to ever tidally circularize as its critical period is @xmath1264.5 days . by extension , the orbit of the eclipsing binary 2mass j0535@xmath170546 ( @xmath127 days ) , which has even lower masses and thus a shorter critical period ,
is also predicted to never circularize , so its nonzero eccentricity is not merely a symptom of its youth ( @xmath1261 myr ; * ? ? ?
@xcite showed that if binary companions are distributed in phase space solely as a function of energy , whether according to the boltzmann distribution or any other arbitrary function , the cumulative distribution of eccentricities must be proportional to @xmath4 and the mean value of @xmath5 should be equal to @xmath128 .
he argued that the observational data for wide stellar binaries at the time was in strong disagreement with this prediction , implying that binaries neither originate from nor relax via dynamical interactions into a state of statistical equilibrium .
tested the @xcite distribution with modern data for solar - type binaries , applying a correction due to eccentricity bias from @xcite , and claimed that the @xmath4 distribution described the wider binaries ( @xmath129 days ) reasonably well .
however , their data still possess a severe lack of high - eccentricity binaries , especially after excluding the three binaries with @xmath1301000-yr orbits , as these happen to include two of the most eccentric orbits in their sample ( @xmath131 ) .
our sample seems to also contravene the @xmath4 distribution , as high-@xmath5 binaries are notably scarce . to quantify this
, we performed ks tests as described in section [ sec : solar ] . following the discussion in section [ sec : vb ] , we corrected for visual binary orbit bias by adding simulated data points uniformly distributed in each of the last two bins ( @xmath5 = 0.60.8 and 0.81.0 ) .
these bins contain 0 and 1 visual binary , respectively , and after accounting for the bias the number in each bin was 0 and 2.4 binaries , averaged over @xmath132 trials .
the resulting probability that the null hypothesis is correct ( i.e. , the observed sample is drawn from the predicted @xmath4 distribution ) was @xmath133 for the entire sample and @xmath134 when considering only the visual binaries in our sample .
because of the potential impact of our assumed bias correction on the results , we also performed ks tests considering only eccentricities below 0.6 , for which there is expected to be negligible observational bias .
these tests gave probabilities of @xmath135 and @xmath136 for the full sample and visual binaries , respectively .
thus , we conclude that our sample of very low - mass binaries does not follow an @xcite distribution over orbital periods of 5.8 to @xmath137 days .
we have also compared our sample to the predictions of numerical simulations that strive to describe the formation of such low - mass binaries .
@xcite simulated the collapse of a 500 molecular cloud , producing sufficiently large numbers of stars and brown dwarfs ( @xmath115 ) to do statistical comparisons with observations .
@xcite reported the parameters of sixteen binaries with total masses of @xmath20 that were generated in this simulation , and we plot their orbital periods and eccentricities in figure [ fig : compare ] .
our observed eccentricity distribution agrees very well with the simulation of @xcite , with a 83% probability of both samples being drawn from the same parent distribution according to a ks test ( 92% for visual binaries alone ) , and this result is not changed if we consider only those simulated binaries within our sample s semimajor axis upper limit of 3.7 au . however , we note that @xcite does not produce any high-@xmath5 ( @xmath6 ) binaries over the same orbital period range as we observe , as all of his high-@xmath5 binaries have very long orbital periods ( @xmath138 days )
. this could be due to the small number statistics of the simulation results , but if not then it is suggestive of a disagreement with the observations that is only made stronger by the fact that the orbit bias correction implies a total of @xmath113 binaries with @xmath139 for our sample
. this discrepancy could be due to the fact that at the end of the simulation the @xcite binaries are still evolving toward shorter orbital periods by dynamical interactions with other protostars and their own accretion disks .
thus , the high-@xmath5 systems could ultimately reach the orbital periods of our sample .
however , their eccentricities would also likely damp down in the process , and thus the discrepancy with observations would remain .
another significant concern is that the very low - mass stars and brown dwarfs in the simulation of @xcite are formed largely by being ejected from their initial gas reservoirs through dynamical interactions .
@xcite showed that by including a more realistic equation of state and radiative feedback , such ejection events become much less common , which resolves the problem that previous simulations ( including @xcite ) produced far too many brown dwarfs .
it remains to be seen how this improvement in the input physics will impact the orbital elements of binaries produced in these simulations . in recent simulations by @xcite ,
very low - mass stars , brown dwarfs , and planetary - mass objects are formed via gravitational instability in the outer parts ( @xmath140 au ) of massive circumstellar disks ( @xmath141 ) .
the orbital periods and eccentricities of the twelve binaries with @xmath19 that were generated in these simulations are shown in figure [ fig : compare ] .
these binaries are a mix of three field binaries ( those that are ejected from the disk ) , and nine binaries that remain bound to the star in a hierarchical triple .
the eccentricities of these two subsets are not statistically different from each other as shown by a ks test ( @xmath142 , 38% chance of being drawn from the same parent distribution ) , and so we simply compare the full set of simulated binaries to our sample .
the simulated binaries have significantly higher eccentricities than we observe , with our ks test giving a probability of only @xmath143 that the two samples were drawn from the same parent distribution .
this result is not changed significantly if we only consider simulated binaries within our sample s semimajor axis upper limit of 3.7 au . as discussed by @xcite ,
the binary components formed in these simulations have their own disks , and after the initial 20 kyr their simulations ignore the gas in order to follow the long term evolution of the system .
the tidal interactions associated with these disks that are not accounted for in the simulations at late times could damp the eccentricities of the simulated binaries .
indeed , @xcite found that including such dissipative interactions was critical in reproducing the stellar eccentricity distribution .
thus , including such additional input physics may solve the disagreement between our observed eccentricities and the simulations of @xcite . on the other hand , such large eccentricities may simply be a hallmark of forming in the frenetic dynamical environment of a fragmenting circumstellar disk .
we have considered for the first time the eccentricity distribution of a large sample of orbits for very low - mass star and brown dwarf binaries .
we have examined potential observational biases in this sample , and for visual binaries we show through monte carlo simulations that by choosing appropriate selection criteria all eccentricities are equally represented ( @xmath2 difference between input and output eccentricity distributions ) .
we find that our sample of very low - mass binaries populate a broad range in eccentricity from nearly circular to highly eccentric ( @xmath144 ) , showing a strong preference for modest eccentricities ( the mean and median of the full sample is @xmath145 ) .
the binaries among our sample that are members of hierarchical triple systems show no special preference for their eccentricity as compared to the other binaries , consistent with a common formation mechanism for both populations .
the overall eccentricity distribution of our sample is similar to solar - type binaries from .
however , very low - mass binaries lack any correlation between eccentricity and period or semimajor axis , which is one of the key features of the solar - type population . in addition , near - circular orbits appear to be quite common among very low - mass binaries , as at least 3 of our 16 sample binaries have @xmath3
. however , such orbits are apparently rare among solar - type binaries ( in the sample , none of the 42 binaries beyond an orbital period of 22 days have @xmath3 ) .
this difference between the two samples is statistically significant according to fisher s exact test ( @xmath96-value of 0.0155 ) .
thus , the orbital data currently available are suggestive of a different formation mechanism or dynamical history for very low - mass binaries as compared to solar - type stars .
we have also tested formation models by comparing our sample of very low - mass binaries to those produced in numerical simulations .
the cluster formation model of @xcite predicts an overall eccentricity distribution that is well - matched to the data .
however , there is a potential discrepancy with the observations due to the lack of any eccentric binaries ( @xmath139 ) at orbital periods corresponding to our sample .
this may be caused by the same physical mechanism that is needed to bring binaries formed in the simulation to these orbital periods , namely tidal damping by the accompanying gas disks .
in contrast , the gravitational instability model of @xcite produces too many high-@xmath5 binaries , with an absence of _ any _ modest eccentricity ( @xmath146 ) orbits .
this may be due to the fact that tidal dissipation by gas disks is not included in the @xcite simulation , causing the resulting eccentricity distribution to be inconsistent with our sample at high significance .
a third possibility is that the observed discrepancies are simply due to multiple formation processes being responsible for producing the field population of very low - mass binaries , so no one model matches all the data .
a further complication is that binaries formed in a cluster environment are likely to undergo additional dynamical processing after their initial eccentricities are set , and thus @xmath147-body simulations are needed to assess the impact of such effects on the final eccentricity distribution . to improve on our observational tests , future efforts should aim to cover a wider range of orbital parameter space , particularly expanding the sample at short orbital periods ( @xmath148 days ) .
the discovery of such tight binaries is only currently feasible from rv monitoring or the `` spectral binary '' technique ( i.e. , identifying the signature of a fainter companion of disparate spectral type in the integrated light spectrum ) .
for example , @xcite used this technique to identify 2mass j0320@xmath170446 as a binary independent of its identification as an sb1 @xcite .
however , even for such spectral binaries the determination of orbits requires either precise rv measurements or resolved astrometry .
this will require new capabilities ( precision near - infrared radial velocities for l and t dwarfs ) as well as new facilities ( e.g. , tmt / gmt / e - elt and _ jwst _ to resolve the much smaller semimajor axes ) .
fortunately , future binaries that will fill in the bottom two orders of magnitude in orbital period will not require the same duration of patient orbital monitoring that has been necessary over the last decade to develop the current sample of @xmath126330-yr period visual orbits .
we have taken the first look at what clues the ensemble of orbital eccentricities of very low - mass stars and brown dwarfs reveal about their formation .
this work follows on previous statistical studies of the visual binary population that found distributions of mass ratio and semimajor axis that were significantly different than for solar - type binaries ( e.g. , * ? ? ?
the theoretical implication that disks may be critical in shaping the eccentricity distribution is in accord with the fact that eccentricity is closely associated with a binary s angular momentum .
this potential relationship is intriguing , as it suggests the possibility for disk - related observational tests .
for example , if high - eccentricity binaries are simply those that lost their disks early in their formation , then perhaps the binary components rotate more rapidly than lower eccentricity binary components due to a lack of disk - braking .
observational evidence has also shown that the formation environment may play an important role in determining the orbits of binaries , with dynamical interactions in the dense orion nebula cluster likely responsible for sculpting the semimajor axis distribution of stellar binaries @xcite while binaries in the sparser taurus region seem to be able to survive at much wider separations @xcite .
further theoretical modeling of the binary formation process taking such effects into account will enable us to better understand the physical processes influencing observed orbital eccentricities .
our research has employed nasa s astrophysical data system ; and the simbad database operated at cds , strasbourg , france .
we are grateful to matthew bate for providing us with the orbital elements of the lowest mass binaries in his simulation .
we have benefitted from discussions with adam kraus about binary formation , john johnson about rv selection effects , and tsevi mazeh about orbital dynamics .
tjd and mcl acknowledge support for this work from nsf grants ast-0507833 and ast-0909222 .
tjd acknowledges support from hubble fellowship grant hst - hf-51271.01-a awarded by the space telescope science institute , which is operated by aura for nasa , under contract nas 5 - 26555 .
finally , the authors wish to recognize and acknowledge the very significant cultural role and reverence that the summit of mauna kea has always had within the indigenous hawaiian community .
we are most fortunate to have the opportunity to conduct observations from this mountain .
_ facilities : _
a general problem in the study of visual binary stars is that orbit determinations are often infeasible , so the true semimajor axis ( @xmath21 ) of the binary must be estimated from the projected separation on the sky ( @xmath38 ) . in the usual approach , @xmath38 is multiplied by a correction factor that accounts for the fact that the observed separation will typically be smaller than the true semimajor axis .
this is always true for circular orbits , but eccentric orbits can be observed at much wider separations than the actual semimajor axis .
@xcite used a correction factor of @xmath149 , computed analytically by assuming edge - on circular orbits .
@xcite performed a monte carlo simulation that assumed random inclinations and reported 1.26 as their time - averaged correction factor .
unfortunately , they did not describe the input eccentricity distribution used nor did they report confidence limits for the correction factor .
@xcite performed several monte carlo simulations assuming random viewing angles and a variety input eccentricity distributions .
he showed that the conversion factor not only varies greatly for different eccentricities , but that the probability distributions for these values are quite broad and can display one or more peaks .
thus , it is important to account for these large uncertainties when converting @xmath38 to @xmath21 . with empirical eccentricity distributions in hand for solar - type and very low - mass binaries
, we have undertaken monte carlo simulations to compute appropriate conversion factors for binaries drawn from these populations .
we assumed uniformly distributed observation times ( @xmath150 ) , uniform arguments of periastron ( @xmath151 ) , and randomly oriented orbits ( @xmath25 ) . from @xmath152 monte carlo trials we computed observed projected separations ( @xmath38 ) and thus arrived at a distribution of conversion factors ( @xmath153 ) .
the results for various assumptions about the input eccentricity distribution are given in table [ tbl : sma - corr ] .
in figure [ fig : sma - corr ] we show the probability distributions of these conversion factors . for comparison , we also show the resulting distribution if @xmath5 is held constant in the simulations . in this case , the distribution has two peaks : one at @xmath154 corresponding to its apastron passage and one at @xmath155 corresponding to periastron .
the empirical distributions of @xmath153 have many small localized peaks due to the fact that they are composed of many individual eccentricity measurements that each contribute two peaks .
the effect is more pronounced for the smaller sample of the thirteen very low - mass binaries as compared to the 46 solar - type binaries .
we have further considered the impact of discovery bias ( see section [ sec : vb ] ) on these distributions .
the resolution limit of an imaging survey will preclude the discovery of binaries when they are at very small separations , which results in truncating the tail of conversion factors at large @xmath153 . in table
[ tbl : sma - corr ] we give the results when considering two cases of discovery bias . in the moderate case , the inner working angle ( iwa ) is a factor of two smaller than the true semimajor axis ( iwa = @xmath156 ) , which truncates the conversion factor distribution at @xmath157 . in the severe case
the resolution limit is equal to the semimajor axis , so only eccentric binaries can be discovered , and the conversion factors are never greater than @xmath158 .
these truncation values are shown as vertical lines in figure [ fig : sma - corr ] .
van leeuwen , f. 2007 , hipparcos , the new reduction of the raw data ( hipparcos , the new reduction of the raw data . by floor van leeuwen , institute of astronomy , cambridge university , cambridge , uk
series : astrophysics and space science library , vol .
350 20 springer dordrecht ) ) was fixed at different multiples of the resolution limit , or inner working angle ( iwa ) .
when the semimajor axis lies inside the iwa highly eccentric orbits are strongly preferred , and as the iwa goes to zero the eccentricity bias disappears .
_ bottom : _ eccentricity bias for binaries with late - m ( solid ) and l dwarf ( dashed ) primaries as determined from monte carlo simulations assuming a magnitude - limited sample with semimajor axis drawn from the log - normal @xcite distribution .
the resolution limit for both simulations was chosen to be the worst - case from among all surveys that discovered the binaries in our sample ( 0@xmath6112 for late - m dwarfs and 0@xmath6107 for l dwarfs ) .
the resulting worst - case eccentricity bias is modest , only slightly ( @xmath115% ) favoring the discovery of highly eccentric ( @xmath64 ) and nearly circular ( @xmath63 ) binaries .
[ fig : bias],width=288 ] ) was fixed at different multiples of the resolution limit , or inner working angle ( iwa ) .
when the semimajor axis lies inside the iwa highly eccentric orbits are strongly preferred , and as the iwa goes to zero the eccentricity bias disappears .
_ bottom : _ eccentricity bias for binaries with late - m ( solid ) and l dwarf ( dashed ) primaries as determined from monte carlo simulations assuming a magnitude - limited sample with semimajor axis drawn from the log - normal @xcite distribution .
the resolution limit for both simulations was chosen to be the worst - case from among all surveys that discovered the binaries in our sample ( 0@xmath6112 for late - m dwarfs and 0@xmath6107 for l dwarfs ) .
the resulting worst - case eccentricity bias is modest , only slightly ( @xmath115% ) favoring the discovery of highly eccentric ( @xmath64 ) and nearly circular ( @xmath63 ) binaries .
[ fig : bias],width=288 ] ) was changed by a factor of two .
the rv observations that discovered ppl 15 only slightly ( @xmath113% ) favored the detection of highly eccentric ( @xmath159 ) binaries .
this effect would be larger ( @xmath1115% ) for binaries with a factor of two lower rv semiamplitudes .
_ bottom : _ same as top panel , except that the dotted colored lines show variations in the eccentricity bias due to having fewer than the three epochs of rv data that were available for the detection of ppl 15 .
surveys with only one or two rv epochs are less sensitive to sb2s at all eccentricities , and they are more biased toward the discovery of highly eccentric binaries .
[ fig : bias - sb2],width=288 ] ) was changed by a factor of two .
the rv observations that discovered ppl 15 only slightly ( @xmath113% ) favored the detection of highly eccentric ( @xmath159 ) binaries .
this effect would be larger ( @xmath1115% ) for binaries with a factor of two lower rv semiamplitudes . _
bottom : _ same as top panel , except that the dotted colored lines show variations in the eccentricity bias due to having fewer than the three epochs of rv data that were available for the detection of ppl 15 .
surveys with only one or two rv epochs are less sensitive to sb2s at all eccentricities , and they are more biased toward the discovery of highly eccentric binaries .
[ fig : bias - sb2],width=288 ] ) .
histograms were generated from @xmath132 randomly drawn eccentricity values for each binary corresponding to the measurement error , so some bins may have a non - integer value .
_ bottom : _ eccentricity plotted as a function of semimajor axis for the visual binary sample ( left ) and as a function of orbital period for the full sample including spectroscopic binaries ( right ) .
triangles denote members of hierarchical triple systems , and circles denote true binaries . [
fig : vlm],title="fig:",width=307 ] 0.2 in ) .
histograms were generated from @xmath132 randomly drawn eccentricity values for each binary corresponding to the measurement error , so some bins may have a non - integer value .
_ bottom : _ eccentricity plotted as a function of semimajor axis for the visual binary sample ( left ) and as a function of orbital period for the full sample including spectroscopic binaries ( right ) .
triangles denote members of hierarchical triple systems , and circles denote true binaries . [
fig : vlm],title="fig:",width=307 ] ) .
histograms were generated from @xmath132 randomly drawn eccentricity values for each binary corresponding to the measurement error , so some bins may have a non - integer value .
_ bottom : _ eccentricity plotted as a function of semimajor axis for the visual binary sample ( left ) and as a function of orbital period for the full sample including spectroscopic binaries ( right ) .
triangles denote members of hierarchical triple systems , and circles denote true binaries . [
fig : vlm],title="fig:",width=307 ] 0.2 in ) .
histograms were generated from @xmath132 randomly drawn eccentricity values for each binary corresponding to the measurement error , so some bins may have a non - integer value .
_ bottom : _ eccentricity plotted as a function of semimajor axis for the visual binary sample ( left ) and as a function of orbital period for the full sample including spectroscopic binaries ( right ) .
triangles denote members of hierarchical triple systems , and circles denote true binaries . [
fig : vlm],title="fig:",width=307 ] 1000-yr orbits excluded .
tidally circularized orbits ( @xmath160 ) are plotted separately in the histogram ( hatched region ) .
triangles and squares denote members of hierarchical triple and quadruple systems , respectively . _
middle : _ cluster formation model simulated binaries from @xcite ; eccentricities for this sample of very low - mass simulated binaries are reported by @xcite . _
bottom : _ circumstellar disk fragmentation model simulated binaries from @xcite , with triangles denoting binaries that remain bound to the simulated host star .
the very low - mass binaries in our sample are shown as filled red symbols on all @xmath96@xmath5 diagrams for comparison .
[ fig : compare],title="fig:",width=307 ] -0.1 in 1000-yr orbits excluded .
tidally circularized orbits ( @xmath160 ) are plotted separately in the histogram ( hatched region ) .
triangles and squares denote members of hierarchical triple and quadruple systems , respectively . _
middle : _ cluster formation model simulated binaries from @xcite ; eccentricities for this sample of very low - mass simulated binaries are reported by @xcite . _
bottom : _ circumstellar disk fragmentation model simulated binaries from @xcite , with triangles denoting binaries that remain bound to the simulated host star .
the very low - mass binaries in our sample are shown as filled red symbols on all @xmath96@xmath5 diagrams for comparison .
[ fig : compare],title="fig:",width=307 ] 1000-yr orbits excluded .
tidally circularized orbits ( @xmath160 ) are plotted separately in the histogram ( hatched region ) .
triangles and squares denote members of hierarchical triple and quadruple systems , respectively . _
middle : _ cluster formation model simulated binaries from @xcite ; eccentricities for this sample of very low - mass simulated binaries are reported by @xcite . _
bottom : _ circumstellar disk fragmentation model simulated binaries from @xcite , with triangles denoting binaries that remain bound to the simulated host star .
the very low - mass binaries in our sample are shown as filled red symbols on all @xmath96@xmath5 diagrams for comparison .
[ fig : compare],title="fig:",width=307 ] -0.1 in 1000-yr orbits excluded .
tidally circularized orbits ( @xmath160 ) are plotted separately in the histogram ( hatched region ) .
triangles and squares denote members of hierarchical triple and quadruple systems , respectively . _
middle : _ cluster formation model simulated binaries from @xcite ; eccentricities for this sample of very low - mass simulated binaries are reported by @xcite .
_ bottom : _ circumstellar disk fragmentation model simulated binaries from @xcite , with triangles denoting binaries that remain bound to the simulated host star .
the very low - mass binaries in our sample are shown as filled red symbols on all @xmath96@xmath5 diagrams for comparison .
[ fig : compare],title="fig:",width=307 ] 1000-yr orbits excluded .
tidally circularized orbits ( @xmath160 ) are plotted separately in the histogram ( hatched region ) .
triangles and squares denote members of hierarchical triple and quadruple systems , respectively . _
middle : _ cluster formation model simulated binaries from @xcite ; eccentricities for this sample of very low - mass simulated binaries are reported by @xcite . _
bottom : _ circumstellar disk fragmentation model simulated binaries from @xcite , with triangles denoting binaries that remain bound to the simulated host star .
the very low - mass binaries in our sample are shown as filled red symbols on all @xmath96@xmath5 diagrams for comparison .
[ fig : compare],title="fig:",width=307 ] -0.1 in 1000-yr orbits excluded .
tidally circularized orbits ( @xmath160 ) are plotted separately in the histogram ( hatched region ) .
triangles and squares denote members of hierarchical triple and quadruple systems , respectively . _
middle : _ cluster formation model simulated binaries from @xcite ; eccentricities for this sample of very low - mass simulated binaries are reported by @xcite . _ bottom : _ circumstellar disk fragmentation model simulated binaries from @xcite , with triangles denoting binaries that remain bound to the simulated host star .
the very low - mass binaries in our sample are shown as filled red symbols on all @xmath96@xmath5 diagrams for comparison .
[ fig : compare],title="fig:",width=307 ] ) to true semimajor axis ( @xmath21 ) for a variety of eccentricity distributions ( see table [ tbl : sma - corr ] ) .
the left panel shows the distribution at an arbitrary fixed eccentricity ( @xmath161 ) , as well as for two analytical eccentricity distributions : uniform and @xmath4 @xcite . for the solar - type binaries ( middle ) and very low - mass binaries ( right ) , we show only the probability distributions for the full samples because subsets of these samples give essentially the same results .
the dotted and dashed lines in the right two panels show where the distributions are truncated in the cases of moderate ( @xmath157 ) and severe ( @xmath158 ) discovery bias , respectively .
[ fig : sma - corr],width=576 ] + 2008 jan 15 & @xmath162 & @xmath163 & & @xmath164 + 2008 sep 8 & @xmath165 & @xmath166 & & @xmath167 + 2009 sep 28 & @xmath168 & @xmath169 & & @xmath170 + 2010 jan 9 & @xmath171 & @xmath172 & & @xmath173 + + + 2010 jan 9 & @xmath174 & @xmath175 & & @xmath176 + 2010 mar 22 & @xmath177 & @xmath178 & @xmath179 & @xmath180 + @xmath96 ( yr ) & @xmath181 & @xmath182 + @xmath21 ( mas ) & @xmath35 & @xmath183 + @xmath5 & @xmath184 & @xmath185 + @xmath186 ( ) & @xmath187 & @xmath188 + @xmath27 ( ) & @xmath189 & @xmath190 + @xmath26 ( ) & @xmath191 & @xmath192 + @xmath193 ( mjd ) & @xmath194 & @xmath195 + ( ) & & @xmath196 + @xmath13 ( dof ) & 6.83 ( 7 ) & 9.38 ( 13 ) + gl 569bab & m8.5 & @xmath197 & 16 & & @xmath198 & @xmath199 & 8 & & 334% & @xmath173% , + 4% & & 0@xmath6105 & 0.52 + 2mass j0920@xmath233517ab & t0 & @xmath200 & 9 & & @xmath201 & @xmath202 & 9 & & 144% & 3% & & 0@xmath6106 & 0.91 + lp 349 - 25ab & m8 & @xmath203 & 10 & & @xmath204 & @xmath205 & 8 & & 76% & 0.6% & & 0@xmath6110 & 0.68 + denis j2252@xmath171730ab & l6 & @xmath206 & 9 & & @xmath207 & @xmath208 & 9 & & 34%56% & 5% & & 0@xmath6107 & 0.48 + hd 130948bc & l4 & @xmath209 & 16 & & @xmath210 & @xmath185 & 1 & & 73% & 0.8% & & 0@xmath6106 & 0.48 + 2mass j2132@xmath231341ab & l5 & @xmath211 & 9 & & @xmath212 & @xmath213 & 9 & & 35% & 5% & & 0@xmath6105 & 0.61 + @xmath214 indi bab & t1 & @xmath215 & 16 & & @xmath216 & @xmath217 & 2 & & 38% & 0.9% & & 0@xmath6108 & 0.12 + lp 415 - 20ab & m7.5 & @xmath218 $ ] & 1 & & @xmath219 $ ] & @xmath184 & 1 & & 55% & @xmath177% , + 9% & & 0@xmath6112 & 1.28 + 2mass j0746@xmath232000ab & l0 & @xmath220 & 5 & & @xmath221 & @xmath222 & 1,11 & & 68% & 2% & & 0@xmath6106 & 0.25 + lhs 2397aab & m8 & @xmath223 & 13 & & @xmath224 & @xmath225 & 7 & & 83% & 3% & & 0@xmath6112 & 0.56 + 2mass j1534@xmath172952ab & t5 & @xmath226 & 15 & & @xmath227 & @xmath228 & 1,11 & & 32%44% & 2% & & 0@xmath6106 & 0.26 + lhs 1070bc & m8.5 & @xmath229 & 3 & & @xmath230 & @xmath231 & 1,14 & & 82% & 0.9% & & 0@xmath6110 & 0.22 + lhs 1901ab & m6.5 & @xmath232 & 12 & & @xmath44 & @xmath233 & 8 & & 39% & 1.4% & & 0@xmath6110 & 0.35 + 2mass j2140@xmath231625ab & m8.5 & @xmath234 $ ] & 1 & & @xmath235 $ ] & @xmath236 & 1,11 & & 38%57% & @xmath1727% , + 53% & & 0@xmath6112 & 0.83 + 2mass j0850@xmath231057ab & l7 & @xmath237 & 17 & & @xmath45 & @xmath238 & 11 & & 10%49% & 100% & & 0@xmath6106 & 0.48 + 2mass j1728@xmath233948ab & l5 & @xmath239 & 17 & & @xmath240 & @xmath241 & 11 & & 20%47% & @xmath1727% , + 167% & & 0@xmath6106 & 0.27 + 2mass j2206@xmath172047ab & m8 & @xmath242 & 4 & & @xmath243 & @xmath244 & 14 & & 20%28% & 2.0% & & 0@xmath6112 & 0.56 + 2mass j1847@xmath235522ab & m6.5 & @xmath245 $ ] & 1 & & @xmath246 $ ] & @xmath247 & 11 & & 9%23% & @xmath1772% , + 194% & & 0@xmath6108 & 0.34 + 2mass j1750@xmath234424ab & m7.5 & @xmath248 $ ] & 1 & & @xmath249 $ ] & @xmath250 & 11 & & 1.3%9% & 60% & & 0@xmath6112 & 0.16 + 2mass j1426@xmath231557ab & m8.5 & @xmath245 $ ] & 1 & & @xmath251 $ ] & @xmath102 & 11 & & 0.2%20% & @xmath17100% , + 73% & & 0@xmath6112 & 0.05 + ppl 15 & m6.5 & s2 & @xmath252 & @xmath253 & @xmath254 $ ] & @xmath255 $ ] & 2 + 2mass j0535@xmath170546 & m6.5 & e & @xmath256 & @xmath257 & @xmath258 & @xmath107 & 12 + 2mass j0320@xmath170446 & m8 & s1 & @xmath259 & @xmath260 & @xmath261 $ ] & [ 0.600.87 ] & 3 + lsr j1610@xmath170040 & d / sdm7 & a , s1 & @xmath262 & @xmath263 & @xmath264 $ ] & [ 0.610.86 ] & 5 + gl 569bab & m8.5 & v & @xmath265 & @xmath199 & @xmath266 & @xmath267 $ ] & 7 + cha h@xmath71 8 & m5.75 & s1 & @xmath268 & @xmath269 & @xmath270 $ ] & [ 0.30.7 ] & 9 + lp 349 - 25ab & m8 & v & @xmath271 & @xmath205 & @xmath272 & @xmath273 $ ] & 7 + hd 130948bc & l4 & v & @xmath274 & @xmath185 & @xmath196 & @xmath275 $ ] & 1 + @xmath214 indi bab & t2.5 & v & @xmath276 & @xmath217 & @xmath277 & @xmath278 $ ] & 4 + 2mass j2132@xmath231341ab & l6 & v & @xmath279 & @xmath213 & @xmath280 $ ] & @xmath281 $ ] & 8 + denis j2252@xmath171730ab & l7.5 & v & @xmath282 & @xmath208 & @xmath283 $ ] & @xmath284 $ ] & 8 + 2mass j0746@xmath232000ab & l0.5 & v & @xmath285 & @xmath222 & @xmath286 & @xmath287 $ ] &
1,10 + lhs 2397aab & m8 & v & @xmath288 & @xmath225 & @xmath289 & @xmath290 $ ] & 6 + lhs 1901ab & m7 & v & @xmath291 & @xmath292 & @xmath233 & @xmath293 $ ] & 7 + lhs 1070bc & m8.5 & v & @xmath294 & @xmath231 & @xmath295 & @xmath296 $ ] & 1,11 + 2mass j1534@xmath172952ab & t5 & v & @xmath297 & @xmath228 & @xmath298 & @xmath299 $ ] & 1,10 + solar - type binaries ( @xmath300 ) & 0.50 & @xmath301 & 0.44 & @xmath302 & 0.57 & @xmath301 + solar - type binaries ( @xmath303 days ) & 0.75 & @xmath304 & 0.82 & @xmath305 & 0.74 & @xmath306 + solar - type binaries ( @xmath129 days ) & 0.25 & @xmath307 & 0.22 & @xmath308 & 0.46 & @xmath309 + @xmath4 @xcite & @xmath133 & @xmath310 & @xmath134 & @xmath311 & @xmath135 & @xmath312 + @xcite & 0.83 & @xmath313 & 0.92 & @xmath314 & 0.64 & @xmath315 + ( * ? ? ?
* @xmath316 au ) & 0.95 & @xmath305 & 0.97 & @xmath317 & 0.91 & @xmath318 + @xcite & @xmath143 & @xmath319 & 0.034 & @xmath320 & 0.19 & @xmath311 + ( * ? ? ?
* @xmath316 au ) & 0.028 & @xmath320 & 0.064 & @xmath321 & 0.43 & @xmath322 + + uniform ( @xmath323 ) & 1.10 & 0.75 , 2.02 & 0.57 , 5.53 + circular ( @xmath120 ) & 1.15 & 1.01 , 1.85 & 1.00 , 4.72 + @xmath4 @xcite & 1.02 & 0.67 , 2.07 & 0.55 , 6.00 + solar - type binaries ( @xmath300 ) & 1.15 & 0.81 , 2.01 & 0.64 , 5.11 + solar - type binaries ( @xmath303 days ) & 1.16 & 0.83 , 2.00 & 0.66 , 5.04 + solar - type binaries ( @xmath129 days ) & 1.14 & 0.78 , 2.02 & 0.61 , 5.29 + very low - mass ( all binaries ) & 1.16 & 0.84 , 1.97 & 0.66 , 5.08 + very low - mass ( visual binaries ) & 1.16 & 0.85 , 1.97 & 0.66 , 5.11 + + + uniform ( @xmath323 ) & 1.02 & 0.72 , 1.46 & 0.57 , 1.89 + circular ( @xmath120 ) & 1.11 & 1.01 , 1.46 & 1.00 , 1.88 + @xmath4 @xcite & 0.91 & 0.65 , 1.42 & 0.55 , 1.88 + solar - type binaries ( @xmath300 ) & 1.06 & 0.79 , 1.51 & 0.63 , 1.90 + solar - type binaries ( @xmath303 days ) & 1.07 & 0.80 , 1.51 & 0.65 , 1.90 + solar - type binaries ( @xmath129 days ) & 1.04 & 0.76 , 1.50 & 0.60 , 1.90 + very low - mass ( all binaries ) & 1.08 & 0.82 , 1.51 & 0.66 , 1.90 + very low - mass ( visual binaries ) & 1.08 & 0.82 , 1.50 & 0.64 , 1.90 + + + uniform ( @xmath323 ) & 0.79 & 0.63 , 0.94 & 0.54 , 0.99 + @xmath4 @xcite & 0.74 & 0.61 , 0.91 & 0.53 , 0.99 + solar - type binaries ( @xmath300 ) & 0.83 & 0.70 , 0.95 & 0.60 , 0.99 + solar - type binaries ( @xmath303 days ) & 0.84 & 0.71 , 0.95 & 0.61 , 0.99 + solar - type binaries ( @xmath129 days ) & 0.82 & 0.68 , 0.94 & 0.58 , 0.99 + very low - mass ( all binaries ) & 0.84 & 0.72 , 0.96 & 0.58 , 0.99 + very low - mass ( visual binaries ) & 0.85 & 0.71 , 0.96 & 0.58 , 0.99 + | we have compiled a sample of 16 orbits for very low - mass stellar ( @xmath0 ) and brown dwarf binaries , including updated orbits for hd 130948bc and lp 415 - 20ab .
this sample enables the first comprehensive study of the eccentricity distribution for such objects .
we find that very low - mass binaries span a broad range of eccentricities from near - circular to highly eccentric ( @xmath1 ) , with a median eccentricity of 0.34 .
we have examined potential observational biases in this sample , and for visual binaries we show through monte carlo simulations that if we choose appropriate selection criteria then all eccentricities are equally represented ( @xmath2 difference between input and output eccentricity distributions ) .
the orbits of this sample of very low - mass binaries show some significant differences from their solar - type counterparts .
they lack a correlation between orbital period and eccentricity and display a much higher fraction of near - circular orbits ( @xmath3 ) than solar - type stars , which together may suggest a different formation mechanism or dynamical history for these two populations .
very low - mass binaries also do not follow the @xmath4 distribution of @xcite , which would be expected if their orbits were distributed in phase space according to a function of energy alone ( e.g. , the boltzmann distribution ) .
we find that current numerical simulations of very low - mass star formation do not completely reproduce the observed properties of our binary sample .
the cluster formation model of @xcite agrees very well with the overall @xmath5 distribution , but the lack of any high-@xmath5 ( @xmath6 ) binaries at orbital periods comparable to our sample suggests that tidal damping due to gas disks may play too large of a role in the simulations .
in contrast , the circumstellar disk fragmentation model of @xcite predicts only high-@xmath5 binaries and thus is highly inconsistent with our sample . these discrepancies could be explained if multiple formation processes are responsible for producing the field population . |
our approach follows the one adopted by scoccimarro et al .
( 2001 ) ; in this framework , the galaxy - galaxy correlation function can be written as + @xmath2 where the first term @xmath3 accounts for pairs of galaxies residing within the same halo , while the @xmath4 represents the contribution coming from galaxies in different haloes .
the above quantities can be written as a function of the mean number of galaxies per halo of mass @xmath5 , @xmath6 , of the spread about this mean value ( also dependent on the mass of the halo hosting the galaxies ) , @xmath7 , of the mean comoving number density of galaxies , @xmath8 ( where @xmath9 is the halo mass function which gives the number density of dark matter haloes per unit mass and volume ) , of the two - point cross - correlation function between haloes of mass @xmath10 and @xmath11 , @xmath12 and finally of the ( spatial ) density distribution of galaxies within the haloes , @xmath13 .
@xmath6 and @xmath7 are the first and second moment of the halo occupation function @xmath14 which gives the probability for a halo of specified mass @xmath5 to contain @xmath15 galaxies .
these can be parameterized as @xmath16 and @xmath17 with @xmath18 in the different mass ranges @xmath19 , @xmath20 and @xmath21 . according to this approach , @xmath22 , @xmath23 , @xmath24 and @xmath25 are parameters to be determined by comparison with observations .
the last ingredient needed for the description of 2-point galaxy clustering is the spatial distribution of galaxies within their haloes .
the first , easiest approach one can take is to assume that galaxies follow the dark matter profile ( navarro , frenk & white 1997 , hereafter nfw ) .
however , since this assumption is not necessarily true , in the present analysis we also consider spatial distributions of the form @xmath26 , with @xmath27 , where the first value corresponds to the singular isothermal sphere case .
all the above profiles have been initially truncated at the virial radius @xmath28 , as one expects galaxies to form within virialized regions , where the overdensity is greater than a certain threshold . however
, this might not be the only possible choice since for instance as a consequence of halo - halo merging
galaxies might also be found in the outer regions of the newly - formed halo , at a distance from the center greater than @xmath28 . in order to perform our analysis ,
we have considered the 4-dimensional grid : @xmath29 ; @xmath30 ; @xmath31 ; @xmath32 .
combinations of these four quantities have then been used to evaluate the mean number density of galaxies @xmath33 . only values for @xmath33 within 2@xmath34 from the observed ones ( derived from the luminosity functions obtained by madgwick et al .
2002 ) were accepted and the corresponding values for @xmath24 , @xmath25 , @xmath22 and @xmath23 have subsequently been plugged into equation ( [ eq : xi ] ) to produce for a specified choice of the distribution profile the predicted galaxy - galaxy correlation function to be compared with the madwick et al .
( 2003 ) results on late - type and early - type galaxies by means of a least squares ( @xmath35 ) fit . 1 .
early - type galaxies ( see figure 1 ) are well described by a halo occupation number of the form broken power - law with @xmath36 , @xmath37 , @xmath38 and @xmath39 , where the two quantities which determine the intermediate - to - high mass behaviour of @xmath40 are measured with a good accuracy .
2 . no model can provide a reasonable fit to the correlation function of late - type galaxies since they all show an excess of power with respect to the data on scales @xmath41\simlt 2 $ ] . in order to obtain an acceptable description of the observations
, one has to assume that star - forming galaxies are distributed within haloes of masses comparable to those of groups and clusters up to two virial radii .
this result is consistent with the phenomenon of morphological segregation whereby late - type galaxies are mostly found in the outer regions of groups or clusters ( extending well beyond their virial radii ) , while passive objects preferentially sink into their centres .
3 . with the above result in mind
, one finds that late - type galaxies ( see figure 2 ) can be described by a halo occupation number of the form single power - law with @xmath42 , @xmath43 and @xmath44 , where the quantities which describe @xmath40 in the high - mass regime are determined with a high degree of accuracy .
4 . within the framework of our models ,
galaxies of any kind seem to follow the underlying distribution of dark matter within haloes as they present the same degree of spatial concentration .
in fact the data indicates both early - type and late - type galaxies to be distributed within their host haloes according to nfw profiles .
we note however that , even though early - type galaxies can also be described by means of a shallower distribution of the form @xmath45 with @xmath46 , this can not be accepted as a fair modelling of the data in the case of late - type galaxies which instead allow for somehow steeper ( @xmath47 profiles . in no case
a @xmath48 density run can provide an acceptable description of the observed correlation function .
these conclusions depend somehow on assuming a specific functional form for the second moment of the halo occupation distribution .
however , magliocchetti & porciani ( 2003 ) have shown that there is not much freedom in the choice of this function if one wants to accurately match the observational data .
an interesting point to note is that results on the spatial distribution of galaxies within haloes and on their halo occupation number are independent from each other .
there is no degeneracy in the determination of @xmath40 and @xmath49 as they dominate the behaviour of the two - point correlation function @xmath50 at different scales .
different distribution profiles in fact principally determine the slope of @xmath50 on small enough ( @xmath51 mpc ) scales which probe the inner regions of the haloes , while the halo occupation number is mainly responsible for the overall normalization of @xmath50 and for its slope on large - to - intermediate scales .
our analysis shows that late - type galaxies can be hosted in haloes with masses smaller than it is the case for early - type objects .
this is probably due to the fact that early - type galaxies are on average more massive ( where the term here refers to stellar mass ) than star - forming objects , especially if one considers the population of irregulars , and points to a relationship between stellar mass of galaxies and mass of the dark matter haloes which host them . 1 madgwick d.s .
( the 2dfgrs team ) , 2002 , mnras , 333 , 133 madgwick d.s . , et al .
( the 2dfgrs team ) , 2003 , mnras , submitted ( astro - ph/0303668 ) magliocchetti m. , porciani c. , 2003 , to appear on mnras , astro - ph/0304003 navarro j.f . , frenk c.s .
, white s.d.m . , 1997 ,
apj , 490 , 493 scoccimarro r. , sheth r.k .
, hui l.,jain b. , 2001 , apj , 546 , 20 | we use the clustering results obtained by madgwick et al .
( 2003 ) for a sample of 96,791 2df galaxies with redshift @xmath0 to study the distribution of late - type and early - type galaxies within dark matter haloes of different mass .
the adopted method relies on the connection between the distribution of sources within haloes and their clustering properties by focussing on the issue of the halo occupation function i.e. the probability distribution of the number of galaxies brighter than some luminosity threshold hosted by a virialized halo of given mass . within this framework
, the distribution of galaxies within haloes is shown to determine galaxy - galaxy clustering on small scales , being responsible for the observed power - law behaviour at separations @xmath1 mpc . for a more extended analysis
, we refer the reader to magliocchetti & porciani ( 2003 ) .
galaxies : clustering - cosmology : theory - cosmology : observations |
there is a strong belief that qcd is equivalent to a string theory and that this equivalence could play an important role in the development of the theory of strong interactions @xcite .
the solution of the two - dimensional qcd in terms of strings has been invoked to put forward the ultimate conjecture that qcd is actually equivalent to string theory .
the @xmath1 string has the property to suppress folding of the surface and therefore has a kind of extrinsic curvature @xcite . in the recent articles @xcite the authors suggested a geometrical string , which directly extends the notion of the feynman integral over paths to an integral over surfaces . in this theory the surface which degenerates into a single particle world line
has an amplitude which is proportional to the length of world line .
this approach was motivated by the fact that in the first approximation the motion of the particle with internal structure is described by the action which should be proportional to the length of the spacetime path @xcite .
this principle , together with the continuity principle allows to define the action of the theory @xmath0 . for the smooth surfaces this action is proportional to the mean extrinsic curvature and
has the property to suppress the folding of the surface @xcite .
the degree of suppression depends on the gonimetric factor @xmath2 in the definition of the action @xmath0 and can be made arbitrary large .
for that the @xmath2 should grow fast enough near the dihedral angles close to @xmath3 which describe a flat surfaces @xcite . in the subsequent publication @xcite
the authors have found that the action @xmath0 has an equivalent representation in terms of the total curvature @xmath4 of the polygons @xmath5 which appear in the intersection of the plane @xmath6 with the given surface @xmath7 .
the curvature @xmath4 should then be integrated over all planes @xmath6 intersecting the surface @xmath7 .
this representation has been proven for the special case of the action @xmath0 , when the action depends linearly on dihedral angle @xmath8 . to provide necessary rigidity of the polygons @xmath9 and the corresponding suppression of the folding of the surface it is important to have generalization of the curvature representation for the cases when the dependence on the dihedral angle @xmath8 is arbitrary @xcite .
the aim of this article is to perform numerical analyses of the curvature representation of the gonihedric action for the cases when dependence on the dihedral angle is arbitrary .
the gonihedric string has been formulated in an euclidean space @xmath10 , where the feynman integral coincides with the partition function for the randomly fluctuating surfaces . in the continuum euclidean
space @xmath10 the random surface is associated with a connected polyhedral surface @xmath7 with vertex coordinates @xmath11 , where @xmath12 and @xmath13 is the number of the vertices .
the action is defined as @xcite @xmath14 where the gonimetric factor @xmath2 is defined as @xmath15 @xmath16 the summation is over all edges @xmath17 of @xmath7 and @xmath18 is a dihedral angle between two neighbor faces ( flat triangles ) of @xmath7 in @xmath10 having a common edge @xmath17 .
the property ( 3 ) always allows to consider the argument of the @xmath2 function to be in the interval @xmath19 in all subsequent computations .
one can say that we are measuring the dihedral angle @xmath18 from that side of the surface where it is less than @xmath3 . as it is easy to see the action indeed has the physical dimension of the length and the degree of the suppression of the folding of the surface depends on @xmath2 @xcite . in cases
when the @xmath2 is equal to @xmath20 and @xmath21 the suppression can be made arbitrary large .
the physical properties of this model and possible connection with qcd have been discussed in @xcite . as we stress above the new representation of action @xmath0
have been proven @xcite for the special case when the factor @xmath2 is linear @xmath22 in the present article we shall perform the numerical analyses of the curvature representation for the cases when the dependence on the dihedral angle is arbitrary .
the numerical integration shows that one can always select the dependence from the dihedral angle so as to provide necessary rigidity of the polygons @xmath23 and therefore corresponding suppression of the folding of the surfaces .
this may help to proceed with the analytical results for the general case .
if we denote by @xmath24 the angle which appears in the intersection of the plane @xmath6 with dihedral angle @xmath18 of the edge @xmath17 , then the total curvature @xmath4 of the polygon or polygons appearing in the intersection is equal to the following sum @xcite @xmath25 here we imply that @xmath24 is equal to @xmath3 for the edges @xmath17 which are not intersected by the given plane @xmath6 . with this definitions the gonihedric action @xmath0 is equal to @xcite @xmath26 where @xmath27 is the invariant measure of two - dimensional planes @xmath6 in @xmath28 @xcite .
this measure can be defined explicitly in @xmath28 through the coordinates of the unit vector @xmath29 which is normal to the plane @xmath6 and through the distance @xmath30 of the plane @xmath6 from the origin @xcite @xmath31 to prove the representation ( 6 ) let us substitute ( 5 ) and ( 7 ) into ( 6 ) , this yields @xcite @xmath32 @xmath33 @xmath34 where @xmath35 now one should prove , that @xmath36 in ( 9 ) , after averaging over all orientations of the plane @xmath6 is indeed equal to @xmath37 ( 4 ) .
this has been proven in @xcite , but here we will present a new proof which allows to apply numerical integration to more complicated cases .
we have the following expression for @xmath38 @xmath39 @xmath40 where @xmath41 is the angle in the intersection of the plane @xmath6 with dihedral angle @xmath8 .
it is very hard to integrate directly ( 9 ) and ( 10 ) to get ( 4 ) . instead of
that we would like to remark , that from ( 9 ) and ( 10 ) it follows that @xmath42 and that @xmath43 where as before we imply that all angles are in the interval @xmath19 ( see property ( 2 ) ) .
the basic property of the dihedral angle @xmath44 is the @xmath45 property @xmath46 which follows from the geometrical picture or one can check it directly using explicit formula ( 10 ) . now substituting ( 13 ) into ( 9 ) and using ( 12 )
we will get @xmath47 the last functional equation together with the boundary condition ( 11 ) has unique solution @xmath48 this completes the proof of the curvature representation ( 6 ) and ( 5 ) for @xmath0 in the particular case ( 4 ) .
our aim is to extend this result into the cases when @xmath36 is an arbitrary function of @xmath37 .
generally one can expect that @xmath49 where the extended curvature @xmath50 is defined as @xmath51 and the function @xmath52 is defined by the integration over all planes @xmath6 @xmath53 now the problem is to select an appropriate function @xmath54 in the definition of the total curvature @xmath50 such that the integration over all planes @xmath6 will produce the gonimetric factor @xmath2 with the necessary properties ( 2 ) and ( 3 ) .
the most important cases are when @xmath54 is a power of @xmath55 @xmath56 or a root function @xmath57 it is also interesting to compute the two angle correlation function @xmath58 which is defined by the integral @xmath59 as it is easy to see from ( 9),(10 ) and from the last integral it is almost impossible to compute these explicitly , therefore we should use the numerical integration to study their behaviour .
in figure 1 we show the result of the integration over all planes @xmath60 in ( 18 ) for various functions @xmath54 .
as we see the qualitative behaviour of the gonimetric function @xmath2 is very similar to the behaviour of the corresponding @xmath54 function .
[ fig31 ] [ fig32 ] therefore it is reasonable to ask how big is the deformation " of the functions ( 19 ) and ( 20 ) after integration over all planes @xmath60 in ( 18 ) , that is the magnitude of the differences @xmath61 or @xmath62 in figure 2 one can see that the differences are not very large , but the equality between these two quantities is definitely excluded . only for @xmath63 does the equality hold @xcite .
the question to which we can not get an answer by numerical analyses is how to select the function @xmath54 in ( 17 ) to recover the power or root behaviour of the @xmath2 function , that is to find out the coefficients @xmath64 in the @xmath54 series @xmath65 such that @xmath66 our numerical results indicate that the biggest coefficient in ( 26 ) is @xmath67 .
figure 3 shows the behaviour of the correlation function @xmath68 for various values of the @xmath69 and @xmath70 . again
the function @xmath68 does not simply coincide with @xmath71
finally we should stress that the curvature representation is very helpful in establishing global properties of the action and hence of the partition function and the loop green functions @xcite . | we analyse the curvature representation of the gonihedric action @xmath0 for the cases when the dependence on the dihedral angle is arbitrary
. = 23.8 cm = 15 cm = -1.2 cm * * |
it has been widely argued that gamma - ray bursts ( grbs ) and afterglows are beamed .
there is much observational evidence that the emission regions of grbs as well as afterglows must be moving relativistically ( for a review see piran 1999 ) .
the emitted radiation is strongly beamed , and one can observe only a region with an opening angle of @xmath0 ( where @xmath1 is the lorentz factor of the emission regions ) off the line of sight . emission outside this narrow cone can not be seen .
these considerations have lead to the analytical speculations and numerical studies on the existence of jets in grbs ( mao & yi 1994 ) and afterglows ( rhoads 1997 , 1999 ; sari , piran & halpern 1999 ; mszros & rees 1999 ; panaitescu & mszros 1999 ; wei & lu 1999 ; moderski , sikora & bulik 1999 ; dermer , chiang & mitman 1999 ; huang et al .
furthermore , the similarity in some of the observed properties between grbs and blazars led to the speculation that jets also appear in grbs ( dermer & chiang 1998 ) .
in addition , jets appear in the context of many models for the grb energy source ( e.g. , ruffert et al . 1997 ; paczyski 1998 ; vietri & stella 1998 ; macfadyen , woosley & heger 1999 ; khokhlov et al . 1999 ; wheeler et al .
1999 ; cheng & dai 1999 ; aloy et al .
1999 ) .
recently , observational evidence for beamed afterglows has been reported .
first , the redshifts of grb 990123 and grb 990510 combined with their observed fluences imply their isotropic energy releases of @xmath2 and @xmath3 ergs respectively ( kulkarni et al . 1999a ; harrison et al .
1999 ) , which are much larger than the energy budgets of the stellar - mass source models .
such enormous energetic budget requires that grb radiation must be highly collimated in some grbs .
second , the observed marked breaks in the light curves of the optical afterglows from these grbs have been interpreted as an observational signature for jets ( kulkarni et al .
1999 ; harrison et al . 1999 ; stanek et al .
1999 ) because , as argued by rhoads ( 1997 , 1999 ) , sari et al .
( 1999 ) , and mszros & rees ( 1999 ) , the lateral expansion and edge effects of a relativistic jet may result in an obvious steepening of the light curve of radiation from the jet , although the other explanation also exists ( dai & lu 1999a , 1999b ) .
third , the observed radio flares of these bursts have been believed to provide an independent and excellent indication of a jetlike geometry in grbs ( kulkarni et al .
1999b ; harrison et al .
as argued by waxman , kulkarni & frail 1998 ) , the radio afterglow from a spherical fireball must rise to peak flux on a timescale of a few weeks ; but owing to the lateral expansion , the radio afterglow from the forward shock of a jet must fade down about few days after the burst ( sari & piran 1999 ) .
therefore , the relative faintness of the observed late - time radio emission implies the existence of a jet .
finally , more importantly , the polarization observed from the afterglows gives further evidence for beamed emission in grbs and afterglows ( see gruzinov 1999 ; sari 1999 ; hjorth et al .
if grbs are similar to bl lacertae objects , then we believe that these two kinds of objects must have similar properties . in this paper , we will discuss the relation between the polarization and the optical magnitude for 4 grbs , which are the only grbs with available data .
here , we summarize the main result of our previous paper ( fan et al . 1997 , 1999 ,
papers i & ii ) .
the observed flux , @xmath4 , of a relativistic jet is related to its intrinsic flux , @xmath5 , by @xmath6 , where @xmath7 , the doppler factor of the jet , is defined by @xmath8^{-1}$ ] , @xmath9 is the velocity in units of the speed of the light , @xmath10 is the lorentz factor , and @xmath11 is the viewing angle .
the value of @xmath12 depends on the shape of the emitted spectrum and the detailed physics of the jet ( lind & blandford 1985 ) , @xmath13 is for a moving sphere and @xmath14 is for the case of a continuous jet , where @xmath15 is the spectral index .
we consider a two - component model , in which the totally observed flux , @xmath16 , is the sum of an unbeamed part @xmath17 and a jet flux , @xmath4 . assuming that the intrinsic flux of the jet , @xmath5 , is some fixed fraction @xmath18 of the unbeamed flux , @xmath17 , i.e. , @xmath19 , we have @xmath20 . if the flux is not totally polarized in the jet , and it is not unreasonable to assume that the jet flux consists of two parts , namely , the polarized and the unpolarized , with the two parts being proportion to each other , i.e. , @xmath21 , @xmath22 , where @xmath23 is a coefficient which determines the polarization of the emission in the jet , then the observed optical polarization can be expressed as @xmath24 where intrinsic polarization , @xmath25 is defined by @xmath26 and @xmath27 is the optical doppler factor . from relation
( 1 ) , we can obtain following relation @xmath28 which follows @xmath29 for the observations of any two epochs , where @xmath30 is the observed magnitude and @xmath31 is the proportional constant . as mentioned above , the @xmath7 can be estimated from @xmath32 , therefore the ratio , @xmath33 can be expressed as @xmath34 if we assume that neither @xmath35 or @xmath17 is changed and @xmath36 is greater than @xmath37 . in this sense
, the ratio @xmath38 can be written in the form of @xmath39 , where the parameter @xmath40 can be expected to be near unity since @xmath41 is usually much greater than @xmath17 .
so , the ratio of polarizations ( 4 ) yields @xmath42 based on a 37-bl lacertae - object sample , @xmath40 is found to be @xmath43 with a chance of probability of @xmath44 ( fan , cheng , zhang 1999 , paper ii ) . in this sense ,
relation ( 4 ) can be expressed in the form @xmath45 where @xmath46 is a constant .
since grbs likely show beaming as in blazar , then similar relation should be expected to exist for grbs .
recently , polarization was measured from the afterglow of gamma ray bursts ( grbs ) : @xmath47 for grb 990510 ( covino et al .
1999 ; wijers et al.1999 ) ; @xmath48 for grb 990123 ( hjorth et al .
1999 ) ; @xmath49 for grb 980329 ( taylor et al . 1998 ) ; @xmath50 for grb 980703 ( frail et al .
these measurements likely suggest that the gamma - ray bursts are beamed ( gruzinov 1999 ; hjorth et al .
1999 ; rhoads 1999 ; panaitescu & meszaros 1998 ; sari 1999 ; sari et al . 1999 ) .
the optical afterglow has been observed from many grbs and found to be variable ( see bloom et al .
1998 ; castrio - tirado et al .
1999a , b ; djorgorski et al . 1997 ; galama et al .
1999 ; garcia et al . 1998 ; groot et al . 1998 ; halpern et al . 1999 ; kulkarni et al . 1999 ; sahu et al . 1997 ; van paradijs et al . 1997 and references therein ) .
since most optical data are made in r band , here we adopted the r magnitude , which is obtained quasi - simultaneously with the measurements of the polarization .
the relevant data are listed in table 1 , where col .
1 gives that name of the grb , col .
2 and 3 the polarization measurements and the corresponding reference ; col . 4 and 5
the photometric data and the corresponding reference . [ cols="<,^,^,^,^ " , ] + notes to the table + @xmath51 polarization obtained in radio band .
+ b98 : bloom et al .
( 1998 ) ; c99 : castro - tirado et al .
( 1999b ) ; co99 : covino et al .
( 1999 ) ; + f98 : frail et al .
( 1998 ) ; h99 : hjorth et al .
( 1999 ) ; + t98 : taylor et al .
( 1998 ) ; w99 : wijers et al .
( 1999 ) the relevant data are shown in fig . 1 , in which the solid line represent @xmath52 .
it is clear that the data of grbs are consistent with the result ( the slope ) obtained for bl lacertae objects ( paper ii ) .
relativistic fireball models can explain the main properties of the grbs and their afterglows ( rees & meszaros 1992 ; paczynski & rhoads 1993 ; katz 1994 ; meszaros & rees 1997 ; vietri 1997 ; waxman 1997 ; sari et al .
1998 ; wijers & galama 1999 ) .
polarization can be observed if the fireball is beamed ( loeb & perna 1998 ; waxman & gruzinov 1999 ; hjorth et al .
the polarization data of bl lacertae objects indicates that the radio selected bl lacertae objects ( rbls ) have very high polarization @xmath54 and x - ray selected bl lacertae objects ( xbls ) also have moderately high polarization , generally @xmath55 . in our previous papers ( fan et al .
1997 , 1999 , paper i & ii ) , we found that the polarization of these bl lacertae objects is associated with the doppler factor and the flux density , particularly , there is a correlation between the observed polarization and the optical magnitude .
when we used the relation to the polarization of grbs , they are consistent with the relation of bl lacertae objects and this suggests that ( 1 ) grbs and bl lacertae objects share similar properties and ( 2 ) the observed polarization in the afterglow of grbs indicate that the emission is beamed .
sari ( 1999 ) and ghisellini & lazzati ( 1999 ) found that , the polarization value , assuming a highly beamed geometry , almost linearly increases with time on a timescale of 1 day after grb although two to three peaks appear in the whole evolution stage of the polarization .
our taken data are about on 1 day after grb ( grb 990123 and grb 990510 ) , so the statistical result presented here is consistent with their theoretical one near this time .
it has been shown that high polarization up to several tens percents is expected for grbs if they are highly beamed ( loeb & perna 1998 ; medvedev & loeb 1999 ; waxman & gruzinov 1999 ; gruzinov 1999 ) .
however , the observed polarization of grb 990510 is only 1.6% .
apparently such low polarization value seems to indicate weak beaming factor .
but there two factors to affect the polarization , namely beaming factor ( boosting factor ) and the orderly magnetic field in the shock front ( intrinsic factor ) . since the optical polarization from bl lacertae objects is emitted from regions much closer to the central power house , typically @xmath56 cm , where the magnetic field is much orderly . on the other hand , the optical polarization of grb afterglows is emitted from regions much farther away from the sources where magnetic field is produced by turbulence and hence should be more disoriented .
in fact , it we assume that the minimum beaming factor of grb 990510 is order of the unity and @xmath57 = 19.1 , then @xmath58 , which is quite consistent with the fireball models .
since only one source , grb 990510 shows definite polarization , the constant , @xmath59 , in the relation @xmath60 is obtained based on the data of grb 990510 .
if more polarization measurements are available , we would expect that the statistical coefficient of @xmath61 is the same , @xmath62 but the constant @xmath46 may be different .
we are grateful for the invaluable comments and suggestions from dr . a. mitra .
this work is partially supported by the rgc grant of hong kong , the national pan deng project of china and the national natural scientific foundation of china .
aloy m. a. et al . 1999 , astro - ph/9910466 castro - tirado a.j .
1999a , apj , 511 , l85 castro - tirado a.j .
1999b , sci , 283 , 2069 cheng k. s. & dai z. g. 1999 , astro - ph/9908248 covino x. et al . 1999 ,
a&al , 348 , l1 dai z. g. & lu t. 1999a , apj , 519 , l155 dai z. g. & lu t. 1999b , astro - ph/9906109 dar a. 1998 , apj , 500 , l93 dermer c. d. & chiang , j. 1998 , astro - ph/9810222 dermer c. d. , chiang , j. & mitman k. e. 1999 , astro - ph/9910240 djorgorski s.g .
1997 , nat .
387 , 876 fan j.h . ,
cheng k.s . , zhang l. , liu c.h .
1997 , a&a , 327 , 947 ( paper i ) fan j.h . ,
cheng k.s .
, zhang l. , 1999 , a&a , ( submitted ) ( paper ii ) frail d.a . , et al . , 1998 , + http://gcn.gsfc.nasa.gov/gcn/gcn3/147.gcn3 galama t.j .
et al . 1999 ,
398 , 394 garcia m.r .
1998 , apj , 500 , l105 ghisellini g. & lazzati d. 1999 , mnras , 309 , l7 groot p.j .
1998 , apj , 502 , l123 gruzinov , a. 1999 , apjl , 525 , l29 harrison f.a .
et al . 1999 , @xmath63 huang y. f. , gou l. j. , dai z. g. & lu t. 1999 , astro - ph/9910 hjorth j. , bjornsson g. , andersen m.i .
et al . 1999 ,
283 , 2073 israel g.l .
et al . 1999 ,
a&a , 348 , l5 katz j.i .
, 1994 , apj , 432 , l107 khokhlov a. m. et al . 1999 ,
apj , in press ( astro - ph/9904419 ) kulkarni s.r .
1999a , nat .
398 , 389 kulkarni s.r .
1999b , apj , 522 , l97 lind k.r . &
blandford r.d .
1985 , apj , 295 , 358 loeb a. , & perna r. 1998 , apj , 495 , 597 macfadyen a. , woosley s. e. & heger a. 1999 , astro - ph/9910034 mao s. d. , & yi i. 1994 , apj , 424 , l131 medvedev m. , & loeb , a. , 1999 , apj , submitted(astro - ph/9904363 ) mszros p. & rees m. r. 1997 , apj , 476 , 232 mszros p. & rees m. r. 1999 , mnras , 306 , 39 moderski r. , sikora m. & bulik t. 1999 , apj , in press ( astro - ph/9904310 ) paczyski b. & rhoads j. , 1993 , apj , 418 , l5 paczyski b. 1999 , apj , 494 , l45 panaitescu a. & mszros p. 1999 , apj , 526 , in press ( astro - ph/9806016 ) piran t. 1999 , phys .
314 , 575 rees m. j. & meszaros p. , 1992 ,
mnras , 258 , l41 rhoads , j. e. 1997 , apj , 485 , l57 rhoads , j. e. 1999 , apj , in press ( astro - ph/9903399 ) sari r. 1999 , apjl , 524 , l43 sari r. , piran t. , 1998 , apjl , 517 , l109 sari r. , piran t. , narayan r. , 1998 , apjl , 497 , l17 sari r. , piran t. , halpern j. , 1999 , apjl , 519 , l17 sahu k.c .
1997 , nat , 387 , 476 stanek k. z. et al . 1999 , apj , 522 , l39 taylor g.b .
1998 , apj 502 , l115 vietri m. 1997 , apj , 478 , l9 vietri m. & stella l. 1998 , apj , 507 , l45 van paradijs j. et al .
1997 , nat . , 387 , 476 waxman e. , 1997 , apj , 485 , l5 waxman , e. , & gruzinov , a. , 1999 , apj , 511 , 852 waxman e. , kulkarni s.r . & frail d.a .
1998 , apj , 497 , 288 wei d. m. & lu t. 1999 , astro - ph/9908273 wheeler j. c. , yi i. , hflich p. & wang l. 1999 , astro - ph/9909293 wijers r.a.m.j .
1999 , apj , 523 , l33 | both observations and theoretical models suggest that the emissions in gamma - ray bursts ( grbs ) and the afterglows are beamed .
we argue that the recent polarization measured in the afterglows gives further evidence of beaming in grbs . in this approach
, we adopted the polarization - magnitude relation of bl lacertae objects to 4 grbs with available polarization measurements and found that the data of the 4 grbs are consistent with the relation of bl lacertae objects .
this result suggests that the emissions in grbs are beamed and there are similarities between grbs and bl lacertae objects ( or generally blazars ) .
we suggest that the weak polarization of grbs may result from low percentage of orderly magnetic field in the shock front . |
this article provides an update on the shell - model interpretation of @xmath0-ray transitions in p - shell hypernuclei @xcite from a previous special issue on recent advances in strangeness nuclear physics and the start of an extension to sd - shell hypernuclei .
the experimental data available at the time was reviewed by tamura in the same volume @xcite and consisted of 22 @xmath0-ray transitions in , , , , , , and , together with a limit on the ground - state doublet spacing in . since then , new results on and from kek e566 using an upgraded germanium detector array , hyperball2 , have been reported at the hyp - x conference by tamura @xcite and ma @xcite .
the ground - state doublet spacing in is established as 161 kev both from the direct observation of the 161 kev @xmath0-ray and from transitions from an excited @xmath5 state at 2832 kev .
the ground - state doublet spacing is closely related to that of which is @xmath6 kev .
it seems that the difference between and can be explained only by invoking the coupling between @xmath1 and @xmath2 hypernuclear states @xcite ( @xmath1-@xmath2 coupling ) .
section [ sec : shell ] describes the shell model calculations and section [ sec : doublet ] summarizes the previously obtained results for transitions observed in the hyperball experiments
. sections [ sec : a8 ] , [ sec : lli9 ] , and [ sec : a10 ] discuss the @xmath7 , 9 , and 10 hypernuclei . section [ sec : be ] contains some information on contributions to the ground - state binding energies of p - shell hypernuclei .
section [ sec : sd ] is devoted to while section [ sec : discussion ] contains a concluding discussion .
shell - model calculations for p - shell hypernuclei start with the hamiltonian @xmath8 where @xmath9 is an empirical hamiltonian for the p - shell core , the single - particle @xmath10 supplies the @xmath11mev mass difference between @xmath1 and @xmath2 , and @xmath12 is the @xmath13 interaction .
the shell - model basis states are chosen to be of the form @xmath14 , where the hyperon is coupled in angular momentum and isospin to eigenstates of the p - shell hamiltonian for the core , with up to three values of @xmath15 contributing for @xmath2-hypernuclear states .
this is known as a weak - coupling basis and , indeed , the mixing of basis states in the hypernuclear eigenstates is generally very small . in this basis ,
the core energies are taken from experiment where possible and from the p - shell calculation otherwise .
the @xmath16 effective interaction can be written @xcite @xmath17 where @xmath18 .
the spin - orbit interactions can alternatively be expressed in terms of the symmetric ( sls ) and antisymmetric ( als ) spin - orbit operators @xmath19 .
the five @xmath20 two - body matrix elements depend on the radial integrals associated with each component in eq .
( [ eq : vlam ] ) , conventionally denoted by the parameters @xmath21 , @xmath22 , @xmath23 , @xmath24 and @xmath25 @xcite . by convention @xcite , @xmath23 and @xmath24 are actually the coefficients of @xmath26 and @xmath27 .
then , the operators associated with @xmath22 and @xmath23 are @xmath28 and @xmath29 . the parametrization of eq .
( [ eq : vlam ] ) applies to the direct @xmath16 interaction , the @xmath16@xmath30 coupling interaction , and the direct @xmath30 interaction for both isospin 1/2 and 3/2 .
values for the parameters based on various nijmegen models of the @xmath13 interactions are given in section 3 of ref .
formally , one could include an overall factor @xmath31 in the analog of eq .
( [ eq : vlam ] ) that defines the interaction , where @xmath32 is the operator that converts a @xmath1 into a @xmath2 .
then , the core operator associated with @xmath33 is @xmath34 .
this leads to a non - zero matrix element only between @xmath1 and @xmath2 states that have the same core , with the value @xmath35 in analogy to fermi @xmath36 decay of the core nucleus .
similarly , the spin - spin term involves @xmath37 for the core and connects core states that have large gamow - teller ( gt ) matrix elements between them .
this point has been emphasized by umeya and harada @xcite in a recent article on the effects of @xmath1-@xmath2 coupling in @xmath38li . in an ls basis for the core ,
the matrix elements of @xmath28 are diagonal ( similarly for @xmath39 ) and depend just on the intensities of the total @xmath40 and @xmath41 for the hypernucleus . because supermultiplet symmetry @xmath42k_cl_cs_cj_ct_c$ ] is generally a good symmetry for p - shell core states , only one or two values of @xmath40 and @xmath41 are important .
the mixing of different @xmath42l_cs_c$ ] is primarily due to the vector ( sls plus als ) terms in the p - shell hamiltonian . of the remaining @xmath16 parameters
, @xmath21 contributes only to the overall binding energy ; @xmath24 does not contribute to doublet splittings in the weak - coupling limit but a negative @xmath24 augments the nuclear spin - orbit interaction and contributes to the spacings between states based on different core states ; in general , there are not simple expressions for the coefficients of @xmath25 .
many hypernuclear calculations have used the venerable cohen and kurath interactions @xcite . here , the p - shell interaction has been refined using the following strategy .
the one - body spin - orbit splitting between the @xmath43 and @xmath44 orbits is fixed to give a good description of the light p - shell nuclei ( say for @xmath45 ) .
the overall strength of the tensor interaction is also fixed , ultimately to produce the cancellation in @xmath46c @xmath36 decay .
the well - determined linear combinations of the central and vector p - shell interactions are then chosen by fitting the energies of a large number of states that are known to be dominantly p - shell in character , including the large spin - orbit splitting at @xmath47 . a detailed discussion of p - shell nuclei is given in section 5 of ref .
@xcite .
@l@ccrrrrrrr & @xmath48 & @xmath49 & @xmath50 & @xmath22 & @xmath23 & @xmath24 & @xmath25 & @xmath51 & @xmath52 + & @xmath53 & @xmath54 & 72 & 628 & @xmath55 & @xmath56 & @xmath57 & 693 & 692 + & @xmath58 & @xmath59 & 74 & 557 & @xmath60 & @xmath61 & @xmath62 & 494 & 471 + & @xmath53 & @xmath59 & @xmath61 & @xmath63 & @xmath64 & @xmath65 & @xmath66 & @xmath67 & 43 + & @xmath58 & @xmath59 & 56 & 339 & @xmath68 & @xmath69 & @xmath70 & 267 & 264 + & @xmath53 & @xmath54 & 61 & 424 & @xmath71 & @xmath72 & @xmath69 & 475 & 505 + & @xmath73 & @xmath5 & 61 & 175 & @xmath74 & @xmath75 & @xmath76 & 153 & 161 + & @xmath77 & @xmath78 & 65 & 451 & @xmath79 & @xmath80 & @xmath69 & 507 & 481 + & @xmath5 & @xmath81 & @xmath82 & @xmath83 & @xmath84 & 1 & 188 & 23 & 26 + & @xmath73 & @xmath85 & 92 & 207 & @xmath86 & 1 & @xmath87 & 248 & 224 +
table [ tab : spacings ] gives a summary of the contributions from @xmath1-@xmath2 coupling and the @xmath16 interaction parameters to all 9 of the measured doublet spacings .
details , such as figures showing spectra and tables giving breakdowns of energy - level spacings , wave functions , and transition rates , can be found in refs .
the set of parameters used for and ( chosen to fit the energy spacings in ) is ( parameters in mev ) @xmath88 the matrix elements for the @xmath1-@xmath2 coupling interaction , based on the g - matrix calculations of ref .
@xcite for the nsc97@xmath89 interactions @xcite , are @xcite @xmath90 these parameters are kept fixed throughout the p - shell in the present calculations .
the ground - state doublet in and the exited - state doublets in and are based on core states that are largely @xmath91 and @xmath92 .
this limits contributions from other than @xmath22 , which enters with a large coefficient of 3/2 in the @xmath91 limit , and @xmath1-@xmath2 coupling .
the coefficient of @xmath22 for the excited state doublet is 7/6 for @xmath93 , @xmath92 , @xmath94 but here @xmath23 and @xmath25 enter with substantial coefficients .
the @xmath95 core state is the lowest member of an @xmath93 , @xmath92 triplet and moves down in energy if the strength of the nuclear spin - orbit interaction is increased , as it is by @xmath24 , in the hypernucleus .
the value of @xmath24 in eq .
( [ eq : param7 ] ) is chosen to reproduce the 2.05-mev excitation energy of the @xmath59 state in .
the @xmath96be core state for the doublet has mainly @xmath93 and @xmath97 , giving a coefficient of @xmath98 for @xmath23 . the contributions from @xmath22 and @xmath25 , arising from small @xmath92 components , together with the small @xmath1-@xmath2 coupling contribution , happen to more or less cancel
this means that the doublet spacing demands a small value for @xmath23 .
the doublet spacings for the heavier p - shell hypernuclei consistently require a smaller value for @xmath22 @xmath99 @xmath25 plays a particularly important role in the ground - state doublet ( @xmath100 ) splitting of and is determined from a measurement of the doublet spacing @xcite .
the ground - state doublet spacing of , which is closely related to that of , is missing form table [ tab : spacings ] because the transition to the @xmath54 state from the 2268-kev @xmath101 level is not observed due to a subtle cancellation @xcite .
however , mesonic weak - decay studies have determined that the ground - state spin - parity of is @xmath53 @xcite . in eq .
( [ eq : param11 ] ) , @xmath24 fits the increase in the excitation energy of the excited - state doublet over the spacing of the p - hole states in @xmath102o . as can be seen from table [ tab : spacings ] , there is a consistent description of the doublet spacings once a larger value of @xmath22 is taken for .
a conjecture , as yet unproven , is that shell - model admixtures beyond @xmath103 for the lightest p - shell nuclei ( @xmath104li in particular ) involve mainly excitations from the s - shell to the p - shell , thus permitting an active role for @xmath105 matrix elements that are larger than those for @xmath106 . for @xmath107 and beyond , higher
admixtures involve @xmath108 excitations and bring in smaller @xmath16 matrix elements . finally , it is clear @xcite that a term such as @xmath24 is necessary to describe the spacings between states based on different core states .
formally , the @xmath24 term arises from a combination of the sls and als interactions but , in practice , @xmath24 is treated as a fitting parameter .
a two - body @xmath109 als interaction that gives rise to similar effects comes from the double one - pion exchange @xmath110 interaction averaged over the @xmath111 wave function , as in the original work of gal , soper , and dalitz @xcite .
while a one - body @xmath24 term appears to be adequate near the beginning and end of the p - shell , there is a need for a much larger effect for ( at least ) , , and .
the high excitation energy ( 1483 kev ) of the first @xmath54 state in , taken together with the known spacings of the ground - state and first - excited state doublets , means that the spacing of the doublet centroids is 1669 kev compared with the @xmath112/@xmath95 core separation of 718 kev .
however , the @xmath24 value of eq .
( [ eq : param11 ] ) gives just over 400 kev towards the difference . while there is sensitivity to the core wave functions , the high excitation energies of 2832 kev for the @xmath85 state of and 4880 kev for the @xmath113 state in
can not be explained either with this value of @xmath24 .
li core states are shown on the left .
the @xmath0-ray branching ratios and lifetimes are theoretical . for each state of
, the calculated energy shifts due to @xmath1-@xmath2 coupling are given .
all energies are in kev.,width=453 ] figure [ fig : lli8 ] gives a theoretical spectrum , including @xmath0-ray branching ratios and lifetimes , for . because the lowest @xmath114 and @xmath115 states of @xmath116li are closely spaced and both have @xmath117 , the bound @xmath5 states of involve significant mixing ( @xmath118 ) of the configurations based on them . on a historical note ,
the spin - parity of @xcite and restrictions on the mixing @xcite were derived from emulsion data on the decay @xmath119 + @xmath120 .
the next largest admixtures are actually from the corresponding @xmath2-hypernuclear states . as can be see from fig .
[ fig : lli8 ] , the energy shift due @xmath1-@xmath2 coupling for the @xmath5 ground state is large in strong contrast to that for the @xmath73 member of the doublet .
in fact , @xmath1-@xmath2 coupling is predicted to account for a third of the doublet spacing .
the predicted energy spacing is very close to the 442.1(21 ) kev energy of a @xmath0-ray observed following the production of highly - excited states of via the @xmath121 reaction @xcite .
the 442 kev @xmath0-ray was tentatively attributed to or its mirror hypernucleus @xcite , the @xmath73 state of which could be reached by @xmath122 deuteron emission from the @xmath123 component of the @xmath95 @xmath124 substitutional state in ( the excited - state doublet in was studied following @xmath125 @xmath126he emission from the same state @xcite ) .
if confirmed , the 442-kev @xmath0-ray would provide additional strong support for the important role played by @xmath1-@xmath2 coupling in hypernuclear spectra .
a 1.22(4)-mev @xmath0-ray , seen after @xmath127 mesons were stopped in a @xmath128be target @xcite , was tentatively ascribed to , possibly as a transition between the @xmath5 states in fig .
[ fig : lli8 ] . in ref .
@xcite , it was found to be difficult to explain such a high energy .
this is still the case despite the addition of 136 kev to the transition energy from @xmath1-@xmath2 coupling .
the role of @xmath1-@xmath2 coupling for the ground - state doublet and the excited @xmath5 state can be seen from the interplay of @xmath33 and @xmath129 in the coupling matrix elements and the use of perturbation theory for the energy shifts ( @xmath130 with @xmath131 mev ) . from eq .
( [ eq : paramls ] ) , @xmath33 gives a contribution of 1.45 mev to the diagonal coupling matrix elements involving the same core state .
adding the remaining @xmath1-@xmath2 contributions ( mainly from @xmath129 ) for the @xmath114 core states gives 2.6524 mev for @xmath132 and 0.7286 mev for @xmath133 .
the off - diagonal matrix elements ( from @xmath129 ) involving @xmath115 and @xmath114 core states are both @xmath134 mev , while the diagonal @xmath115 matrix element is 1.4761 mev . in the weak - coupling limit ,
the push on the ground state is 88 kev from the @xmath135 state and 28 kev from the @xmath136 state while the push on the @xmath73 state is only 7 kev .
the mixing between the @xmath137 and @xmath138 states increases the @xmath1-@xmath2 coupling matrix elements for the lower @xmath5 state and decreases them for the upper state ; putting in the numbers from above , an estimate of 157 kev is obtained for the push on the ground state .
as far as electromagnetic transitions are concerned , the p - shell wave functions account well for the m1 properties of the @xmath114 and @xmath115 core states using the bare m1 operator , leaving room for the expected small enhancement of the isovector matrix elements by meson - exchange currents @xcite ; the calculated magnetic moments of @xmath116li and @xmath116be are 3.145 @xmath139 and @xmath140 @xmath139 compared with the experimental values of 3.256 @xmath139 and @xmath141 @xmath139 , respectively . because the m1 matrix element for the ground - state doublet transition is proportional to @xmath142 ( @xmath143 ) in the weak - coupling limit @xcite the transition is going to be much faster in the odd - proton nucleus .
this simple approximation does not apply because of the configuration mixing in the @xmath5 wave functions .
the m1 transition strengths end up being 1.086 w.u . in and
0.043 w.u . in ( @xmath144 ps ) . for the decay of the @xmath85 state in ,
one obtains a branch of 64% to the ground state and a lifetime of 46 fs .
all these low - energy transitions are little affected by their e2 components . as a further example of the effects of @xmath1-@xmath2 coupling , the ground - state doublet
is predicted to have a spacing of 101 kev with @xmath1-@xmath2 contributions of 154 and 182 kev to the binding energies of the @xmath5 and @xmath73 states , respectively . in this case , @xmath1-@xmath2 coupling reduces the doublet spacing by 28 kev .
there is interest in because it has been studied using the @xmath128be@xmath4 reaction at jlab @xcite and could be studied via the @xmath128be@xmath3 reaction at j - parc .
in addition , it is the possible source of a 1303-kev @xmath0-ray seen in a stopped @xmath127 experiment @xcite , most strongly on a @xmath128be target .
figure [ fig : lli9 ] gives a theoretical spectrum , including @xmath0-ray branching ratios and lifetimes , for .
because non - spin - flip production is dominant in the @xmath145 reaction at rest , the only likely candidate for the 1303-kev @xmath0-ray is the excited @xmath53 to ground state transition . in this case
, an 840-kev transition to the lowest @xmath59 state should also be observable .
the predicted energy of 1430 kev is too high . for comparison , the @xmath16 parameter set in eq .
( [ eq : param11 ] ) predicts the @xmath77 state at 1331 kev and the @xmath146 state at 471 kev .
it is clear that an in - flight @xmath3 study with the hyperball - j at an incident @xmath127 energy where spin - flip amplitudes are important is desirable .
li core states are shown on the left along with the spectroscopic factors for proton removal from @xmath128be .
the @xmath0-ray branching ratios and lifetimes are theoretical . for each state of
, the calculated energy shifts due to @xmath1-@xmath2 coupling are given .
all energies are in kev . on the right , the structural factors ( defined in appendix a ) giving the relative population of levels in purely non - spin - flip ( @xmath147 ) and purely spin - flip ( @xmath148 ) production reactions on a @xmath128be target are given.,width=453 ] from the @xmath128be@xmath149li study by liu and fortune @xcite and the pickup spectroscopic factors given in fig .
[ fig : lli9 ] , the bulk of the cross section for the production of @xmath111 states in the @xmath128be@xmath4 reaction is expected to be concentrated in states built on the lowest three states of @xmath96li ( see ref .
@xcite for an early theoretical study ) .
this is indeed the case @xcite .
the strongest observed state is the upper member of the ground - state doublet but there appears to be more strength in the states based on the @xmath95 core state than predicted and a disagreement about the location of the strength based on the @xmath112 state .
the strength based on the @xmath150 state is close enough to the neutron threshold that the states should be narrow and any significant strength associated with them should be observable in the electro - production reaction
. this would be the case for the ( 8 - 16)2bme and ( 8 - 16)pot interactions of cohen and kurath @xcite but not for the ( 6 - 16)2bme interactions or the various fitted interactions used in recent hypernuclear studies .
the former interactions favor the second @xmath112 state in proton removal from @xmath128be because the lowest @xmath112 state is dominantly @xmath117 , @xmath92 rather than strongly mixed @xmath97 and @xmath92 . the use of the ( 8 - 16)2bme interaction is the reason that the spectrum of umeya and harada @xcite looks rather different from the one in fig .
[ fig : lli9 ] .
figure [ fig : lbe10 ] gives a theoretical spectrum , including @xmath0-ray branching ratios , pickup spectroscopic factors , and formation factors for .
is another hypernucleus that could be studied via the @xmath3 reaction with the hyperball - j at j - parc , this time with a @xmath151b target .
a strong reason for doing so would be to try to measure the ground - state doublet spacing by observing transitions to both members from a higher level .
the obvious candidate is the @xmath73 level based on the @xmath152 core level of @xmath128be .
unfortunately , the @xmath153 transition is strongly hindered with respect to the @xmath154 transition by a factor of 15 from the recoupling coefficient ( but gains something back on the @xmath155 factor ) . in the weak - coupling limit
, the @xmath153 branch would be only 9% but something is gained from configuration mixing .
again , the e2 components of the transitions are not very important . the spacings of the ground - state and excited - state doublets
are predicted to be very similar .
this could certainly be checked in a @xmath3 experiment and which @xmath0-ray is which could be determined by choosing @xmath127 momenta for which the ratio of spin - flip to non - spin - flip is quite different .
the @xmath5 level based on the broad @xmath115 level in @xmath128be could be populated via the reaction on @xmath151be if only a thick enough @xmath151be target could be made .
be core states are shown on the left along with the spectroscopic factors for proton removal from @xmath151b .
the @xmath0-ray branching ratios and lifetimes are theoretical . for each state of
, the calculated energy shifts due to @xmath1-@xmath2 coupling are given .
all energies are in kev . on the right , the structural factors giving the relative population of levels in purely non - spin - flip ( @xmath147 ) and purely spin - flip ( @xmath148 ) production reactions on a @xmath151b target are given.,width=453 ] as far
as production reactions are concerned , the @xmath151b reaction has been studied in kek e336 @xcite . from the spectroscopic factors for proton or neutron removal from @xmath151b ,
one expects to see four strong peaks up to about 10 mev in excitation energy . in the experiment ,
these are not cleanly resolved but the data has been fitted to extract energies and relative yields @xcite .
the relative yields are in good agreement with the @xmath147 structure factors in fig .
[ fig : lbe10 ] but the spectrum is somewhat expanded with respect to that extracted from the data ( the extracted energy may be affected by the steeply rising background that extends under the fourth peak ) . a spectrum in which the four peaks , based on the core states reached strongly by proton removal from @xmath151b , are cleanly separated has recently been obtained using the @xmath151b@xmath4 reaction at jlab @xcite .
this is as predicted by motoba et al .
@xcite and by the results in fig .
[ fig : lbe10 ] based on more recent information on in - medium @xmath13 interactions . states involving a @xmath156 coupled to the same core states are also expected to be strongly populated ( again see ref .
@xcite ) and it will be interesting to make a detailed comparison between theory and experiment .
@c@rrrrrrrrr & @xmath157 & @xmath1-@xmath2 & @xmath22 & @xmath23 & @xmath24 & @xmath25 & sum & @xmath158 & @xmath159 + & @xmath54 & 101 & 1 & 0 & 176 & 0 & 278 & & + & @xmath5 & 154 & 152 & @xmath75 & 454 & @xmath160 & 709 & 7.16(70 ) & @xmath161 + & @xmath54 & 253 & 6 & 0 & 619 & 1 & 879 & & + & @xmath54 & 78 & 419 & 0 & 94 & @xmath79 & 589 & 5.58(3 ) & @xmath162 + & @xmath5 & 160 & 288 & @xmath163 & 192 & @xmath57 & 625 & 6.80(3 ) & @xmath164 + & @xmath53 & 183 & 350 & @xmath69 & 434 & @xmath163 & 952 & 8.50(12 ) & @xmath165 + & @xmath5 & 275 & 175 & @xmath166 & 595 & @xmath74 & 1022 & & + & @xmath54 & 4 & 0 & 0 & 207 & 0 & 211 & 6.71(4 ) & + & @xmath54 & 99 & 2 & 0 & 540 & 0 & 641 & & + & @xmath81 & 158 & @xmath167 & @xmath168 & 554 & 127 & 748 & & + & @xmath5 & 35 & 125 & @xmath75 & 386 & @xmath168 & 518 & 8.89(12 ) & @xmath169 + & @xmath59 & 66 & 203 & @xmath84 & 652 & @xmath170 & 858 & 10.24(5 ) & @xmath171 + & @xmath5 & 103 & 108 & @xmath63 & 704 & @xmath172 & 869 & 11.37(6 ) & @xmath169 + & @xmath54 & 130 & 197 & @xmath163 & 621 & @xmath173 & 885 & & + & @xmath5 & 255 & 115 & @xmath75 & 458 & @xmath174 & 785 & & + & @xmath54 & 28 & @xmath56 & 0 & 841 & @xmath55 & 864 & 11.69(12 ) & @xmath175 + & @xmath5 & 75 & 47 & 6 & 816 & @xmath176 & 904 & 12.17(33 ) & @xmath177 + & @xmath54 & 116 & 8 & 0 & 636 & 2 & 762 & & + & @xmath53 & 59 & 40 & 12 & 630 & @xmath178 & 726 & 13.59(15 ) & @xmath179 + & @xmath5 & 62 & 94 & 6 & 349 & @xmath180 & 412 & 13.76(16 ) & @xmath181 + table [ tab : be ] shows the @xmath1-@xmath2 and spin - dependent contributions to the ground - state binding energies for a wide range of p - shell hypernuclei .
the sum of these contributions can reach 1 mev .
the experimental @xmath182 values are from emulsion studies @xcite ( a number of mirror hypernuclei are not listed ) except for where the value comes from a study using the @xmath183o@xmath4 reaction @xcite ( the observed @xmath5 state is actually 26 kev above the @xmath81 ground state @xcite ) .
the remaining hypernuclei that are listed are chosen because the neutron - rich p - shell cores have higher isospin and can exhibit larger effects from @xmath1-@xmath2 coupling .
the spin - independent central component @xmath21 of the @xmath16 interaction does nt affect the spectra but can be estimated from the binding energies by taking @xmath182 ( ) = 3.12 mev as the @xmath111 single - particle energy and using @xmath184 where @xmath185 is the number of p - shell nucleons in the core .
the values of @xmath21 so extracted are given in the last column of table [ tab : be ] .
the @xmath21 are relatively constant and close to the values derived from @xmath16 potential models @xcite . in reality , small repulsive contributions quadratic in @xmath185 are expected from the double one - pion exchange interaction @xcite which would then call for a somewhat more attractive @xmath21 .
quite good estimates can be made for @xmath182 values . in the case of @xmath186 hypernuclei , two spin - averaged @xmath182 values enter into the binding energy along with a @xmath186 two - body matrix element that is known to be quite small ( @xmath187 mev ) .
then , it is clear that the knowledge of single-@xmath1 binding energies can be used to make reliable estimates for the binding energies of @xmath186 hypernuclei @xcite .
an extension of the studies of @xmath0-ray transitions in p - shell hypernuclei is planned for in j - parc e13 @xcite .
the reason for choosing @xmath188f as a target is that @xmath189f has a primarily @xmath190 , @xmath191 ground state so that one should observe a relatively large ground - state doublet spacing for , in complete analogy to ( and the doublet based on the second @xmath112 state in @xmath102n ) ; @xmath188f itself has a primarily @xmath190 , @xmath192 ground state , which is why an ( impractical ) @xmath193ne target was considered by millener et al .
@xcite .
the @xmath189f core nucleus has quite a dense spectrum , including @xmath194 , @xmath195 , @xmath196 , and @xmath197 states close to 1 mev ( see fig.3 in ref .
the wave functions for the lowest @xmath112 and @xmath95 states are given in a @xmath198-coupling basis in table [ tab : cw ] .
the @xmath112 state is actually 92.7% @xmath190 , @xmath191 ( amplitudes 0.8985 and @xmath199 for su(3 ) symmetry @xmath200 and @xmath201 , respectively ) , while the @xmath95 state is 96.9% @xmath202 , @xmath191 ( amplitudes 0.9722 and @xmath203 ) . historically , this simplicity for @xmath189f and @xmath188f was a significant factor in the introduction of elliott s su(3 ) model @xcite .
@c@rrrrrr @xmath157 & @xmath204 & @xmath205 & @xmath206 & @xmath207 & @xmath208 & @xmath209 + @xmath112 & 0.6038 & @xmath210 & @xmath211 & & 0.1130 & 0.4386 + @xmath95 & 0.5721 & @xmath212 & @xmath213 & 0.7844 & & + @c@rrrrr & & ho & ws & ws & ws + & @xmath214 & @xmath215 fm & @xmath216 & @xmath217 mev & @xmath218 mev + @xmath219 & 0 & @xmath220 & @xmath221 & @xmath222 & @xmath223 + & 1 & @xmath224 & @xmath225 & @xmath226 & @xmath227 + @xmath228 & 1 & @xmath229 & @xmath230 & @xmath231 & @xmath232 + @xmath233 & 1 & @xmath234 & @xmath235 & @xmath236 & @xmath237 + & 2 & @xmath238 & @xmath239 & @xmath240 & @xmath241 + @xmath242 & 2 & @xmath243 & @xmath244 & @xmath245 & @xmath246 + @xmath247 & 2 & @xmath248 & @xmath249 & @xmath250 & @xmath251 + & 3 & @xmath252 & @xmath253 & @xmath254 & @xmath255 + in the sd - shell , there are 8 @xmath256 matrix elements ; 4 central , one each for ls and als in relative p states , and 2 tensor ( in both even and odd states ) .
these are shown as a function of binding energy in table [ tab : sd ] . here , a radial representation of the @xmath16 interaction that reproduces the matrix elements of eq.([eq : param11 ] ) is used ( cf . ref .
table [ tab : sd ] demonstrates that the matrix elements are sensitive to the binding energies of the sd - shell orbits , especially the noded @xmath257 orbit .
we note that sd - shell orbits indeed become loosely bound and the @xmath257 orbit moves below the @xmath258 orbit for states in p - shell hypernuclei . combining the woods - saxon matrix elements for the exp .
case in table [ tab : sd ] with the wave functions in table [ tab : cw ] , the doublet spacings for states based on the lowest @xmath112 and @xmath95 states are 305 kev and 196 kev , respectively .
this calculation is for simple weak - coupling states without the inclusion of @xmath1-@xmath2 coupling and calculations similar to those performed for p - shell hypernuclei remain to be performed . in addition , @xmath189f has low - lying negative parity states that can be reached in the reaction @xcite .
the lowest @xmath81 and @xmath5 states are predominantly of the form @xmath259 ( 89% and 81% for @xmath81 and @xmath5 in a full @xmath260 shell - model calculation ; alternatively 72.4% and 74.6% @xmath261 su(3 ) symmetry ) .
the @xmath260 hypernuclear basis requires @xmath262 states in addition to @xmath263 and @xmath264 to make a non - spurious basis .
calculations using refined ( @xmath103 ) interactions for p - shell core nuclei , with the inclusion of @xmath1-@xmath2 coupling , have been quite successful in that a large body of data on hypernuclear level spacings has been correlated with relatively few @xmath13 parameters .
the introduction of explicit @xmath1-@xmath2 coupling is generally beneficial .
it is necessary to understand the s - shell hypernuclear binding energies , especially for and .
it makes significant contributions of varying size relative to the dominant @xmath16 spin - spin interaction in p - shell doublet spacings and binding energies .
in particular , it seems necessary to understand the ground - state doublet spacings in both ( a limit ) and .
these effects involve the interplay of fermi and gamow - teller type matrix elements connecting core states .
the remaining problems are to understand ( 1 ) the need for different @xmath16 spin - spin interaction strengths at either end of the p - shell and ( 2 ) the need for a stronger enhancement of the nuclear vector interaction terms ( ls and als ) near mid shell by the presence of the @xmath1 .
the next steps are to expand the shell - model basis and to reintroduce the double one - pion exchange @xmath110 interaction considered by gal , soper , and dalitz @xcite . for consistency
, one has to go beyond @xmath265 states for the core and the @xmath1 configurations which makes for a challenging problem .
a somewhat more tractable problem is to treat the full @xmath260 basis of hypernuclear states @xcite .
this is necessary to ( 1 ) treat properly @xmath156 states in both the p- and sd - shells and ( 2 ) estimate decay widths for particle emission from unbound hypernuclear states @xcite , this being the way in which @xmath0 transitions in daughter hypernuclei have been studied .
this work has been supported by the us department of energy under contract no .
de - ac02 - 98ch10886 with brookhaven national laboratory .
for a particular @xmath266 transition , it is possible to pull out a structure factor that multiplies a particular distorted ( or plane ) wave radial integral and governs the relative cross sections for related states .
see , for example , section 3.2 of ref .
@xcite where the structure factor is @xmath267 times the square of an ls one - body density - matrix ( obdme ) for the transition , together with some common factors such as the square of the isospin clebsch - gordan coefficient for the transition . for the predominantly non - spin - flip transitions in or reactions
, there is a single obdme with @xmath268 . in the case
@xmath4 reactions , the kroll - ruderman term @xmath269 is dominant and one needs to evaluate the magnetization current contributions to the transverse electric and magnetic operators that appear in the @xmath270 cross section , specifically the @xmath2 and @xmath271 terms that appear in eqs.(22b ) and ( 22c ) , and given in eqs.(1d ) and ( 1e ) , of donnelly and haxton @xcite . for the same l , we can pull out a common radial factor , basically the longitudinal form factor @xmath272 . for the electric terms with @xmath273 , we get just @xmath272 , while for the magnetic terms we get @xmath274 for @xmath275 and @xmath276 for @xmath277 . to get the structure factors , we multiply by the obdme with the given ( lsj ) , square , and add the statistical @xmath267 factor . note that in the case of a simple particle - hole excitation for a closed - shell target nucleus , the @xmath198 obdme is just a phase factor so that the ( lsj ) obdme is given by a normalized 9j symbol for the @xmath278 transformation . in this paper
, we have just the simple @xmath279 transition so that for @xmath268 we need the obdme @xmath280 , while for @xmath191 we need the combination @xmath281 .
these are multiplied by @xmath282 , where @xmath283 is the isospin clebsch - gordan coefficient . | the interpretation of hypernuclear @xmath0-ray data for p - shell hypernuclei in terms of shell - model calculations that include the coupling of @xmath1- and @xmath2-hypernuclear states is briefly reviewed .
next , / and are considered , both to exhibit features of @xmath1-@xmath2 coupling and as possible source of observed , but unassigned , hypernuclear @xmath0 rays .
then , the feasibility of measuring the ground - state doublet spacing of be , which , like , could be studied via the @xmath3 reaction , is investigated .
structural information relevant to the population of states in these hypernuclei in recent @xmath4 studies is also given .
finally , the extension of the shell - model calculations to sd - shell hypernuclei is briefly considered .
hypernuclei ; shell - model |
a. wilmshurst @xcite showed that there is an upper bound on the number of zeros of a harmonic polynomial @xmath5 , where @xmath6 and @xmath7 are analytic polynomials of different degree , answering the question of t. sheil - small @xcite .
let @xmath8 .
wilmshurst showed that @xmath9 is a sharp upper bound when @xmath10 and conjectured that the upper bound is actually @xmath11 .
d. khavinson and g. wiatek @xcite showed that wilmshurst s conjecture holds for the case @xmath12 using methods from complex dynamics . when hearing of this result , p. poggi - corradini asked whether this approach can be extended to the case @xmath13 , where @xmath6 and @xmath7 are analytic polynomials . in this note
, we apply the approach from @xcite to prove [ thm : mainresult ] let @xmath14 be a rational function where @xmath6 and @xmath7 are relatively prime , analytic polynomials and such that @xmath15 .
then @xmath16 we note that the zeros of @xmath0 are isolated , because each zero is also a fixed point of @xmath17 , an analytic rational function of degree @xmath9 .
this also follows from a result of p. davis @xcite ( chapter 14 ) concerning the schwarz functions of analytic curves .
( the schwarz function is an analytic function @xmath18 that gives the equation of a curve in the form @xmath19 , _ cf . _
a rational schwarz function implies that the curve is a line or a circle , so the degree must be one .
we also note that @xmath0 will not have a zero at @xmath20 . if @xmath20 were a zero , then @xmath21 . by the change of variable @xmath22
, we see that @xmath23 has a zero at @xmath24 . restricting @xmath25 to the real axis
, we see that @xmath26 and @xmath27 .
we obtain a contradiction when we restrict @xmath25 to the imaginary axis .
we now discuss an application of our result to gravitational microlensing .
an @xmath4-point gravitational lens can be modeled as follows .
suppose that we have @xmath4 point masses ( such as stars ) .
construct a plane through the center of mass of these point masses , such that the line of sight from the observer to the center of mass is orthogonal to this plane .
this plane is called the lens plane ( or deflector plane ) .
suppose that the lens plane is between the observer and a light source .
( we are assuming that the distance between the point masses is small compared to the distance between the observer and the lens plane , as well as the distance between the lens plane and the light source . ) the plane containing our light source which is parallel to the lens plane is called the source plane . due to the deflection of light by the point masses ,
multiple images of the light source are formed .
this phenomenon is known as gravitational microlensing .
see @xcite and @xcite for an introduction to gravitational lensing and @xcite for an introduction to a complex formulation of lensing theory ; also see @xcite .
gravitational microlensing can be modeled by a lens equation , which defines a mapping from the lens plane to the source plane . to set up a lens equation for our @xmath4-point gravitational lens , the point masses
are projected onto the lens plane .
the projection of the @xmath28-th point mass has a scaled mass of @xmath29 and is located at a scaled coordinate of @xmath30 in the lens plane , where @xmath29 is a positive constant and @xmath30 is a complex constant .
suppose that we have a light source located at a scaled coordinate of @xmath25 in the source plane .
following @xcite , this lens equation will be given by @xmath31sign@xmath32 , where the normalized ( external ) shear @xmath33 and the optical depth ( or normalized surface density ) @xmath34 are real constants .
in this model , if @xmath35 satisfies the lens equation , then our gravitational lens will map @xmath35 to @xmath25 ; hence @xmath35 corresponds to the position of a lensed image .
the number of lensed images is the number of solutions of the lens equation .
we can rewrite this lens equation in terms of the rational harmonic function @xmath36 by letting @xmath37sign@xmath38 .
we thus see that the zeros of @xmath39 are solutions of the lens equation for a light source at position @xmath25 .
h. witt @xcite showed for @xmath2 that the maximum number of observed images is at most @xmath40 when @xmath41 and @xmath42 when @xmath43 .
a. petters @xcite used morse theory to obtain further estimates for both cases .
s. mao , a. petters , and h. witt @xcite conjectured that the maximum number of images is linear in @xmath4 for the case @xmath41 and @xmath44
. s. h. rhie @xcite later conjectured that for @xmath2 such a gravitational lens gives at most @xmath45 images . in the @xmath41 case , @xmath46 and
theorem 1 settles this conjecture .
further , for the case @xmath43 , we see that @xmath47 , so theorem 1 gives an upper bound of @xmath48 lensed images .
as a result , we have the following corollary : an @xmath4-point gravitational lens modeled by @xmath31sign@xmath32 , where @xmath2 can produce at most @xmath3 images when the shear @xmath41 and at most @xmath49 images when the shear @xmath43 .
the authors are indebted to professor jeffrey rabin of the university of california , san diego , for pointing out that our result settles a conjecture in gravitational lensing , for alerting us to the papers of s. h. rhie ( @xcite and @xcite ) , and for noting that @xcite addresses the question of sharpness .
we wish to thank professor donald sarason of the university of california , berkeley for helpful discussions concerning rhie s sharpness proof .
the second author is grateful to professor larry weaver of kansas state university for helpful discussions concerning gravitational lensing .
both authors wish to thank professor shude mao of the jodrell bank observatory at the university of manchester and professor arlie petters of duke university for valuable comments on the second draft of this paper .
we first recall some terminology and a fact from complex dynamics .
let @xmath14 be a rational function , where @xmath6 and @xmath7 are relatively prime , analytic polynomials .
the degree of @xmath50 , denoted by @xmath51 , is given by @xmath52 .
@xmath35 is called a _ critical point of _ @xmath50 if @xmath53
. we will be interested in counting the fixed points of a rational function .
a fixed point @xmath54 is said to be _ attractive _ , _ repelling _ or _ neutral _ if @xmath55 , @xmath56 , or @xmath57 respectively . a neutral fixed point where the derivative is a root of unity
is said to be _ rationally neutral_. a fixed point @xmath58 is said to _ attract _ some point @xmath59 if the iterates of @xmath50 at @xmath25 converge to @xmath58 .
we will be using the following result , whose proof can be found in @xcite , chapter iii , theorems 2.2 and 2.3 : [ fact : fatou ] let @xmath50 be a rational function with @xmath60 . if @xmath58 is an attracting or rationally neutral fixed point , then @xmath58 attracts some critical point of @xmath50 .
we will also use a version of the argument principle for harmonic functions .
let @xmath61 be harmonic in an open set .
we say that @xmath35 is in the _ critical set _ of @xmath61 if the jacobian of @xmath61 vanishes at @xmath35 .
in particular , in a sufficiently small disk containing @xmath35 , there exist analytic functions @xmath62 and @xmath63 such that @xmath64 in this disk .
then @xmath35 is in the critical set of @xmath61 if @xmath65 .
we say that @xmath35 is a _ singular zero _ of @xmath61 if @xmath35 is in the critical set of @xmath61 and @xmath66 .
@xmath61 is said to be _ sense - preserving _ at @xmath35 if @xmath67 and is said to be _ sense - reversing _ at @xmath35 if @xmath68 .
the _ order _ of a non - singular zero is positive if @xmath61 is sense - preserving at the zero and negative if @xmath61 is sense - reversing .
let @xmath61 be harmonic in a punctured neighborhood of @xmath58 .
we will refer to @xmath58 as a _ pole _ of @xmath61 if @xmath69 as @xmath70 .
let @xmath71 be an oriented closed curve that contains neither zeros nor poles of @xmath61 .
the notation @xmath72 denotes the increment in the argument of @xmath39 along @xmath71 . following @xcite ,
the _ order _ of a pole of @xmath61 is given by @xmath73 , where @xmath71 is a sufficiently small circle around the pole .
we note that if @xmath61 is sense - reversing on a sufficiently small circle around the pole , then the order of the pole will be negative .
we will use the following version of the argument principle which is taken from @xcite : [ fact : argprinc ] let @xmath61 be harmonic , except for a finite number of poles , in a simply connected domain @xmath74 in the complex plane .
let @xmath71 be a jordan curve contained in @xmath74 not passing through a pole or a zero , and let @xmath75 be the open bounded region created by @xmath71 .
suppose that @xmath61 has no singular zeros in @xmath74 and let @xmath76 be the sum of the orders of the zeros of @xmath61 in @xmath75 .
let @xmath77 be the sum of the orders of the poles of @xmath61 in @xmath75
. then @xmath78 .
we note that a more general form of the argument principle can be found in @xcite .
[ prop : sensepreszeros ] let @xmath6 and @xmath7 be relatively prime analytic polynomials . if @xmath79 is a rational function of degree @xmath2 , then the set of points for which @xmath80 and @xmath81 has cardinality at most @xmath82 .
_ let @xmath83 denote the number of points @xmath58 satisfying the conditions of proposition [ prop : sensepreszeros ] . following the approach of d. khavinson and g. wiatek @xcite , we consider the function @xmath17 , which is a rational function of degree @xmath9 .
as in @xcite , we note that all fixed points of @xmath84 that are critical points for @xmath85 are rationally neutral fixed points for @xmath86 , so fact [ fact : fatou ] will apply . their outline ( lemmas 1 - 3 in @xcite together with fact [
fact : fatou ] ) carry over word for word to the rational case if @xmath87 is replaced by the riemann sphere @xmath88 .
in particular , each point @xmath58 which satisfies the conditions of proposition [ prop : sensepreszeros ] attracts at least @xmath89 critical points of @xmath90 . since @xmath91 , by the riemann - hurwitz relation ( see @xcite , section 17.14 ) , @xmath90 has @xmath92 critical points ( counted with multiplicities ) in @xmath88 .
@xmath83 will be largest when all of the critical points of @xmath90 are attracted to points @xmath58 satisfying the conditions of proposition [ prop : sensepreszeros ] . since different fixed points
attract disjoint sets of critical points , we see that @xmath93 and the claim follows .
as in @xcite , we say that the rational function @xmath1 is _ regular _ if no critical point of @xmath85 is a zero of @xmath61 .
this will allow us to apply fact [ fact : argprinc ] to count the zeros of @xmath61 .
[ lem : senserevzeros ] if @xmath50 is regular of degree @xmath2 , then @xmath85 has at most @xmath45 distinct finite zeros . _ proof .
_ let @xmath79 , where @xmath6 and @xmath7 are relatively prime , analytic polynomials .
let @xmath95 and @xmath96 . then @xmath97 . let @xmath98 denote the number of sense - reversing zeros of @xmath61 and let @xmath83 be defined as in the proof of proposition 1 .
since @xmath50 is regular , no zero of @xmath61 lies on the critical set of @xmath61 ; this will allow us to apply fact [ fact : argprinc ] .
in particular , @xmath83 counts the number of sense - preserving zeros of @xmath61 .
then @xmath61 has @xmath99 zeros in the finite complex plane , counting multiplicity .
note that @xmath61 is sense - reversing at each of the @xmath100 poles of @xmath50 ; hence , the order of each pole is negative .
consider the increment in the argument of @xmath61 in a region bounded by a circle of large enough radius so that all of the finite zeros and poles of @xmath61 are enclosed .
we will apply fact [ fact : argprinc ] and use our bounds on @xmath83 from proposition [ prop : sensepreszeros ] . note that we are counting zeros with multiplicity .
we consider two cases . 1 .
assume that @xmath101 .
then the critical set of @xmath61 is bounded .
for @xmath35 large , @xmath1 is at most @xmath102 .
thus @xmath61 is sense - preserving on our circle with an argument change of @xmath103 .
we also note that @xmath20 is not a zero of @xmath61 .
by fact [ fact : argprinc ] , @xmath104 .
hence @xmath105 , so @xmath61 has at most @xmath45 zeros in @xmath88 and the claim follows .
2 . assume that @xmath106 .
since @xmath6 and @xmath7 are relatively prime , at least one of the two polynomials has a non - zero constant term .
+ we first suppose that @xmath6 has a non - zero constant term .
then @xmath107 . let @xmath22 and consider @xmath108 .
then @xmath109 satisfies the conditions of the previous case , replacing @xmath100 by @xmath110 and @xmath111 by @xmath112 .
thus , @xmath109 has at most @xmath45 zeros . since @xmath107
, @xmath61 must also have at most @xmath45 zeros in the finite complex plane .
+ we now suppose that @xmath6 does not have a non - zero constant term .
since @xmath6 and @xmath7 are relatively prime , the lowest order term in @xmath7 must be a non - zero constant .
consider @xmath113 .
for all @xmath114 sufficiently small , we have that @xmath115 and @xmath7 are relatively prime . in that case ,
@xmath61 and @xmath116 have the same poles and @xmath116 approaches @xmath61 uniformly as @xmath117 on compact subsets of @xmath87 which do not contain any of the poles of @xmath61 .
we then substitute @xmath22 to form @xmath118 @xmath119 . by construction , @xmath120 , so @xmath121 has at most @xmath45 zeros in the finite plane .
hence , this also holds for @xmath116 since @xmath122 . +
now suppose that @xmath61 has more than @xmath45 zeros in @xmath87 . by constructing a sufficiently small circle around each zero of @xmath61 , we can guarantee that @xmath61 is harmonic in each of the resulting closed disks .
moreover , we can make our circles sufficiently small so that each closed disk contains no other zeros of @xmath61 and contains no critical points of @xmath61 ( recall that @xmath50 is regular ) .
assume that @xmath116 does not vanish in such a disk surrounding a zero , so @xmath123 .
then , for @xmath114 sufficiently small , @xmath124 has zero increment around the circle bounding this disk , by fact [ fact : argprinc ] ; this contradicts @xmath61 having a zero in this disk .
we see that @xmath116 must have the same number of zeros as @xmath61 in that disk ( counting multiplicity ) .
thus , @xmath116 must have more than @xmath45 zeros in the finite plane , a contradiction . @xmath94
it remains to show that it is enough to consider regular rational functions .
[ lem : regdense ] if @xmath1 is a rational function of degree greater than @xmath125 , then the set of complex numbers @xmath114 for which @xmath126 is regular is open and dense in @xmath87 .
_ as in @xcite , it is enough to show that the image of the critical set of @xmath85 is nowhere dense in @xmath87 .
in contrast to the polynomial case , it is possible for the critical set to be unbounded .
this difficulty however is easily resolved by restricting @xmath61 to an increasing family of concentric balls whose radii run to @xmath20 .
@xmath94 lemma [ lem : regdense ] shows that the set of regular rational functions is dense in the topology of uniform convergence in the spherical metric .
as at the end of the proof of the ansatz , it is easily seen that every zero of the function @xmath85 must be a limit point for zeros of @xmath127 where @xmath117 .
thus , lemma [ lem : regdense ] and the ansatz prove the theorem .
our proof does not indicate whether the @xmath45 bound is always sharp .
we have , however , found an example where this bound is attained for the case @xmath128 ; namely , @xmath129 .
this function has @xmath130 distinct finite zeros .
this function has two poles : @xmath131 and @xmath132 ; @xmath61 is sense - reversing at these poles .
there are three sense - reversing zeros : @xmath133 , @xmath134 , and @xmath135 .
as expected from fact [ fact : argprinc ] for an overall index change of @xmath136 on large circles , there are two sense - preserving zeros ; namely , @xmath137 and @xmath138 , figure [ fig : crit ] is a mathematica plot of the critical set of this function and figure [ fig : imag ] shows its image .
we note that @xmath61 is sense - preserving in the unbounded component in figure [ fig : crit ] , sense - reversing in the larger of the two bounded components , and sense - preserving in the smaller of the two bounded components .
we also note that @xmath61 can not be rewritten to model a 2-point gravitational lens .
l. geyer @xcite has recently shown that the @xmath139 bound on the number of zeros of @xmath140 where @xmath141 is sharp for all @xmath2 .
d. bshouty and a. lyzzaik @xcite have recently given an elementary proof for @xmath142 .
hence , a sharp bound on the number of zeros of @xmath85 must be at least @xmath139 .
results from gravitational lensing help with the question of sharpness of the @xmath45 bound in theorem 1 .
we restrict our attention to the @xmath41 , @xmath44 case of the lens equation ( _ cf . _
introduction ) .
mao , petters , and witt @xcite have shown there are @xmath143 images when the light source is at the origin and the @xmath144 point masses ( each of mass @xmath145 ) are vertices of a regular polygon centered at the origin .
thus , in contrast to the case where @xmath1 is a polynomial , by moving the poles of @xmath1 into the finite plane , we see that a sharp bound on the number of zeros must be at least @xmath143 .
further , by a clever perturbation argument , rhie @xcite has shown that the @xmath45 bound is attained and is hence sharp for all @xmath2 .
it is known that the number of lensed images is odd in the case of gravitational lensing by a regular gravitational lens ( @xcite , @xcite ) , given that the position of the source is not a critical value of the lens equation .
( a regular lens has a smooth mass distribution .
a point @xmath25 is a critical value of the lens equation if the jacobian of the lens mapping defined by the lens equation vanishes at any of the points in @xmath146 . )
given that @xmath3 is not always odd , it may seem surprising that @xmath45 lensed images may be possible for an @xmath4-point gravitational lens with @xmath41 and @xmath44 . however , it was shown by petters @xcite using morse theory , that this is not a problem .
we summarize this result in the following corollary and present an alternate proof .
suppose that we have an @xmath4-point gravitational lens modeled as above with @xmath41 and a light source which is not located at a critical value of the lens equation
. then the number of lensed images must be even when @xmath4 is odd and odd when @xmath4 is even .
_ for this case of the lens equation , we may apply step ( 1 ) in the proof of the ansatz .
in particular , the number of images @xmath76 will be @xmath147 , where @xmath83 denotes the number of sense - preserving solutions and @xmath98 the number of sense - reversing solutions of the lens equation . from step ( 1 ) , we see that @xmath148 , so @xmath149 .
@xmath94 this result has also been shown by other techniques in @xcite ( footnote 2 , page 2 ) , which extends the approach of w. burke @xcite for the case of a regular gravitational lens to that of an @xmath4-point gravitational lens .
note that the above proof is similar to the approach used by n. straumann @xcite for the case of a regular gravitational lens .
we also note that theorem 1 can be applied to the case of a more general mass distribution than point masses . for a compactly supported mass distribution with the projected mass density @xmath150 in the lens plane
, the lens equation transforms into @xmath151 ( _ cf .
_ @xcite and @xcite ) .
in particular , we have the following corollary : suppose that the projected mass density of a gravitational lens consists of @xmath2 radially symmetric , continuous , compactly supported densities in the lens plane .
if the shear @xmath41 , then the number of lensed images outside the support of the masses can not exceed @xmath3 .
if the shear @xmath43 , then the number of lensed images outside the support of the masses can not exceed @xmath49 .
_ consider the integral term in the lens equation for one of the mass densities .
a calculation shows that this integral evaluated for any @xmath35 outside of the support of this radially symmetric mass density equals @xmath152 , where @xmath153 is the center of the mass density and @xmath154 is a finite constant .
hence , for @xmath35 outside of the support of the masses , the lens equation reduces to the lens equation for @xmath4 point masses ( each mass is reduced to a point mass at the sphere s center ) .
corollary 1 can then be applied to count the possible images , giving the upper bounds on images outside the support of the masses as claimed .
@xmath94 if the mass distribution is composed entirely of luminous matter , then the only lensed images that will be visible are those that lie outside of the support of the masses . however , there are mass distributions that are not luminous ( see @xcite and @xcite ) .
for such mass distributions , it is possible to have lensed images that lie inside the support of the masses . a question related to bounding the number of zeros of @xmath85 would be to find a bound on the number of zeros of @xmath155 , where both @xmath156 and @xmath50 are rational functions .
it is not clear what condition would be needed for such a function to have a finite number of zeros , much less a bound on the number of distinct zeros ( beyond the obvious bzout theorem s bound when we assume that there are a finite number of zeros ) .
for example , let @xmath157 be a polynomial ( of degree at least two ) and consider @xmath158 . the zeros of this function form a lemniscate in the complex plane .
in other words , @xmath39 has an infinite number of zeros and these zeros are not isolated .
moreover , @xmath159 . if we compare this with the result of wilmshurst @xcite for the case when @xmath61 is entire : @xmath69 as @xmath160 implies a finite number of zeros , we see that all of the zeros for an entire function will be isolated under wilmshurst s condition for finite valence . thus , @xmath61 having finite poles makes the problem of discreteness of its zero set much more subtle .
xxxx 99 nasa astronomy picture of the day , _ giant cluster bends , breaks images - 2001 june 10 _ , http://antwrp.gsfc.nasa.gov/apod/ap010610.html .
nasa astronomy picture of the day , _ dark matter map - 2003 august 14 _ , http://antwrp.gsfc.nasa.gov/apod/ap030814.html .
d. bshouty and a. lyzzaik , _ on crofoot - sarason s conjecture for harmonic polynomials _ , preprint ( 2003 ) .
w. l. burke , _ multiple gravitational imaging by distributed masses _ ,
j. * 244 * ( 1981 ) , l1 .
l. carleson and t. gamelin , _ complex dynamics _ , springer - verlag , new york - berlin - heidelberg ( 1993 ) .
mr 94h:30033 . p.
j. davis , _ the schwarz function and its applications _ , the carus mathematical monographs , no .
17 , the mathematical association of america , buffalo , n. y. , 1974 . mr 53 # 11031 .
o. forster , _ lectures on riemann surfaces _ , translated from the german by bruce gilligan , graduate texts in mathematics , 81 , springer - verlag , new york - berlin ( 1981 ) .
. l. geyer , _
sharp bounds for the valence of certain harmonic polynomials _ , preprint ( 2003 ) .
d. khavinson and g. wiatek , _ on the number of zeros of certain harmonic polynomials _ ,
* 131 * ( 2003 ) , 409 - 414 .
. s. mao , a. o. petters , and h. j. witt , _ properties of point mass lenses on a regular polygon and the problem of maximum number of images _ , ` proceeding of the eighth marcel grossmann meeting on general relativity ( jerusalem , israel , 1997 ) ' , edited by t. piran , world scientific , singapore ( 1998 ) , 1494 - 1496 , arxiv : astro - ph/9708111 .
r. narayan and m. bartelmann , _ lectures on gravitational lensing _ , in ` proceedings of the 1995 jerusalem winter school ' ( 1995 ) ; + online version http://cfa-www.harvard.edu/@xmath161narayan/papers/jerulect.ps .
a. o. petters , _ morse theory and gravitational microlensing _ , j. math . phys .
* 33 * ( 1992 ) , 1915 - 1931 .
mr 93c:58060 .
a. o. petters , h. levine , and j. wambsganss , _ singularity theory and gravitational lensing _ , birkhuser , boston ( 2001 ) .
. s. h. rhie , _ can a gravitational quadruple lens produce 17 images ? _ , arxiv : astro - ph/0103463 . s. h. rhie , _
@xmath4-point gravitational lenses with @xmath162 images _ , arxiv : astro - ph/0305166 . t. sheil - small , _ complex polynomials _ , cambridge studies in advanced mathematics 73 , cambridge university press ( 2002 ) .
mr 2004b:30001 .
t. sheil - small in _ tagesbericht _ , mathematisches forsch .
oberwolfach , funktionentheorie , 16 - 22.2.1992 , 19 .
n. straumann , _ complex formulation of lensing theory and applications _
, helvetica physica acta * 70 * ( 1997 ) , 894 - 908 , arxiv : astro - ph/9703103 . t. j. suffridge and j. w. thompson , _ local behavior of harmonic mappings _
, complex variables theory appl .
* 41 * ( 2000 ) , 63 - 80 .
mr 2001a:30019 .
j. wambsganss , _ gravitational lensing in astronomy _ , living rev .
relativity * 1 * ( 1998 ) [ online article ] : cited on january 26 , 2004 , http://www.livingreviews.org / articles / volume1/1998 - 12wamb/. a. s. wilmshurst , _ the valence of harmonic polynomials _ , proc .
math . soc .
* 126 * ( 1998 ) , 2077 - 2081 .
. h. j. witt , _ investigation of high amplification events in light curves of gravitationally lensed quasars _ ,
. astrophys . * 236 * ( 1990 ) , 311 - 322 . | extending a result from @xcite , we show that the rational harmonic function @xmath0 , where @xmath1 is a rational function of degree @xmath2 , has no more than @xmath3 complex zeros .
applications to gravitational lensing are discussed . in particular
, this result settles a conjecture @xcite concerning the maximum number of lensed images due to an @xmath4-point gravitational lens . |
one of the main open problems in schubert calculus is to find an analog of the littlewood - richardson rule for flag manifolds ( * ? ? ?
* problem 11 ) , and more generally to find combinatorial formulae for intersection numbers of schubert varieties .
this problem was recently solved by coskun for two - step flag manifolds @xcite .
we give such a combinatorial interpretation for intersection numbers of grassmannian schubert problems on any type @xmath0 flag manifold .
this number counts certain objects that we call filtered tableaux which satisfy conditions coming from the schubert problem .
when the flag manifold is a grassmannian this coincides with a standard interpretation of these numbers obtained from the littlewood - richardson rule .
grassmannian schubert problems on the flag manifold were studied in @xcite ; they are exactly the schubert problems which appear in the generalization of the shapiro conjecture to flag manifolds given there . in section [ sec : rule ]
we define filtered tableaux , give an example , and state our formula , which we prove in section [ sec : proofs ] .
our proof uses some identities of @xcite which were established using geometry , and is thus not completely combinatorial . in section [ sec : comments ] we explain how our formula relates to one coming from monk s formula @xcite and discuss how to give a purely combinatorial proof based on the rule of kogan @xcite .
we thank the centre de recherches mathmatiques in montral and the organizers of the workshop on combinatorial hopf algebras and macdonald polynomials in may 2007 where this collaboration was begun .
we also thank hugh thomas for helpful discussions , and chris hillar for comments on the manuscript .
for background on flag manifolds and schubert calculus , see @xcite .
we fix a positive integer @xmath1 throughout .
let @xmath2 be a non - empty subset of @xmath3:=\{1,2,\dotsc , n{-}1\}$ ] , which we write in increasing order @xmath4 a is a sequence @xmath5 of linear subspaces in @xmath6 @xmath7 where @xmath8 .
the set of all flags of type @xmath9 is a complex manifold of dimension @xmath10 schubert varieties and classes in @xmath11 are indexed by permutations @xmath12 of @xmath13 whose descent set is contained in @xmath9 . for a permutation @xmath12 ,
let @xmath14 be the class of the schubert variety corresponding to @xmath12 , following the conventions in @xcite .
its cohomological degree is @xmath15 , where @xmath16 counts the number of inversions @xmath17 of @xmath12 .
if @xmath18 is another subset then there is a projection @xmath19 whose fibres are products of flag varieties .
when @xmath20 is a singleton , @xmath21 is the grassmannian @xmath22 of @xmath23-planes in @xmath6 . in this case
, we write @xmath24 for @xmath25 .
we note that @xmath26 is just the schubert class @xmath27 .
schubert classes in @xmath22 are also indexed by partitions @xmath28 , which are northwest - justified arrays of boxes in a @xmath29 rectangle , @xmath30{figures / rect.eps}}}_b$ ] .
associated to a partition @xmath28 is the @xmath12 with @xmath28 and descent at @xmath23 .
this permutation has a unique descent at @xmath23 , and its first @xmath23 values are @xmath31 here , @xmath32 denotes the number of boxes in row @xmath33 of @xmath28 .
we write @xmath34 for the grassmannian schubert class @xmath14 . here
are three partitions with @xmath35 and @xmath36 ; the third is also drawn inside @xmath30{figures / rect.eps}}}_3 $ ] .
they correspond to the grassmannian permutations @xmath37 , @xmath38 , and @xmath39 .
@xmath40 let @xmath41 be the number of boxes in @xmath28 .
this is half the cohomological degree of the schubert class @xmath34 and is the complex codimension of the associated schubert variety .
the littlewood - richardson rule for the grassmannian expresses a product @xmath42 of two schubert classes as a sum of classes @xmath43 where @xmath44 with @xmath45 . in this rule ,
the coefficient of @xmath43 is the number of of skew shape @xmath46 and content @xmath28 .
these are fillings of the boxes in @xmath47 with positive integers such that 1 .
the entries weakly increase left - to - right across each row and strictly increase down each column . 2 .
the number of @xmath48s in the filling is equal to @xmath49 , the number of boxes in row @xmath48 of @xmath28 .
if we read the entries right - to - left across each row and from the top row to the bottom row , then at every step we will have encountered at least as many occurrences of @xmath33 as of @xmath50 for each positive integer @xmath33 .
for example , here are some littlewood - richardson tableaux . @xmath51{figures / lr1.eps } } \put(3.5,2.5){2 } \put(17,15.5){1 } \put(30,28.5){1 } \end{picture } \quad \begin{picture}(52,52)(0,-6.5 ) \put(0,0){\includegraphics[height=39pt]{figures / lr1.eps } } \put(4,2.5){1 } \put(16.5,15.5){2 } \put(30,28.5){1 } \end{picture } \quad \qquad \begin{picture}(52,52 ) \put(0,0){\includegraphics[height=52pt]{figures / lr2.eps } } \put(43,41.5){1 } \put(29.5,28.5){2 } \put(17 , 15.5){1}\put(29.5,15.5){3 } \put(4,2.5){1}\put(16.5 , 2.5){2 } \end{picture } \qquad \begin{picture}(52,52 ) \put(0,0){\includegraphics[height=52pt]{figures / lr2.eps } } \put(43,41.5){1 } \put(30 , 28.5){1 } \put(16.5,15.5){2}\put(29.5,15.5){2 } \put(4,2.5){1}\put(16.5 , 2.5){3 } \end{picture } \qquad \begin{picture}(52,52 ) \put(0,0){\includegraphics[height=52pt]{figures / lr2.eps } } \put(43,41.5){1 } \put(30 ,
28.5){1 } \put(17 , 15.5){1}\put(29.5,15.5){2 } \put(3.5,2.5){2}\put(16.5 , 2.5){3 } \end{picture } $ } \ ] ] a in the cohomology ring of @xmath11 is the pullback of a schubert class along a projection to a grassmannian .
that is , it has the form @xmath52 where @xmath53 and @xmath54{figures / rect.eps}}}_b$ ] .
these are indexed by pairs @xmath55 with @xmath54{figures / rect.eps}}}_b$ ] .
a is a list @xmath56 with @xmath57 .
we require that for every @xmath58 we have @xmath59 and @xmath60{figures / rect.eps}}}_{a_i}$ ] , and also @xmath61 by the dimension condition , we have @xmath62}_\alpha\,,\ ] ] where @xmath63}_\alpha$ ] is the class of a point in @xmath11 .
the problem that we solve is to give a combinatorial formula for the coefficient of @xmath63}_\alpha$ ] in this product .
note that if @xmath64 this coefficient is zero ( e.g. by ( * ? ? ?
* lemma 1 ) ) , and so we will generally assume that @xmath65 . write @xmath30{figures / shift.eps}}}_\alpha$ ] for the union of all rectangles @xmath30{figures / rect.eps}}}_a$ ] for each @xmath66 , where the rectangles all share the same upper right corner . here
are three such shapes when @xmath36 .
@xmath67{figures / shift.eps}}}_{235}\ = \ $ } \
\includegraphics[height=40pt]{figures / shift235.eps } \quad\qquad \raisebox{30pt}{${{\includegraphics[height=8pt]{figures / shift.eps}}}_{145}\ = \ $ } \ \includegraphics[height=40pt]{figures / shift145.eps}\quad\qquad \raisebox{30pt}{${{\includegraphics[height=8pt]{figures / shift.eps}}}_{[6]}\ = \ $ } \ \includegraphics[height=40pt]{figures / shift6.eps}\ ] ] a @xmath68{figures / shift.eps}}}_\alpha$ ] is a subset of boxes which are northwest justified .
for example , when @xmath69 , the shaded boxes are four shapes in @xmath30{figures / shift.eps}}}_{234}$ ] .
@xmath70{figures / sh1.eps } \quad\qquad \includegraphics[height=40pt]{figures / sh2.eps } \quad\qquad \includegraphics[height=40pt]{figures / sh3.eps } \quad\qquad \includegraphics[height=40pt]{figures / sh4.eps } \quad\qquad\ ] ] let @xmath71 be a grassmannian schubert problem
. set @xmath72 and fix a shape @xmath68{figures / shift.eps}}}_\alpha$ ] .
a is a sequence @xmath73 of shapes together with fillings @xmath74 of the skew shapes @xmath75 by positive integers which satisfy the following properties . 1 . the skew shape @xmath75 must fit entirely within the rectangle @xmath30{figures / rect.eps}}}_{a_i}\subset{{\includegraphics[height=8pt]{figures / shift.eps}}}_\alpha$ ] .
2 . the filling @xmath76 is a littlewood - richardson tableau of content @xmath77 .
note that we must have @xmath78 .
an induction shows that the coefficient of @xmath63}_b=\sigma_{\includegraphics[height=6pt]{figures / rect.eps}_b}$ ] in a product @xmath79 in @xmath80 is the number of filtered tableaux with shape @xmath30{figures / rect.eps}}}_b$ ] whose content is the sequence @xmath81 .
we generalize this to any flag manifold .
[ thm : main ] let @xmath71 be a grassmannian schubert problem on @xmath11 . then the coefficient of @xmath63}_\alpha$ ] in the product @xmath82 is the number of filtered tableaux with shape @xmath30{figures / shift.eps}}}_\alpha$ ] and content @xmath83 .
we use this formula to compute the intersection number @xmath84 , defined by @xmath85}_\alpha\ = \
\pi_1^*(\sigma_{\includegraphics[height=4pt]{figures/3.1.eps}})\cdot \pi_2^*(\sigma_{\includegraphics[height=4pt]{figures/3.2.eps}})\cdot \pi_3^*(\sigma_{\includegraphics[height=8pt]{figures/21.3.eps}})\cdot \pi_4^*(\sigma_{\includegraphics[height=12pt]{figures/111.4.eps}})\cdot \pi_5^*(\sigma_{\includegraphics[height=12pt]{figures/111.5.eps}})\,.\ ] ] here , @xmath86 $ ] and @xmath30{figures / shift.eps}}}_\alpha$ ] is the full staircase shape .
there are exactly three sequences of shapes @xmath87 which satisfy the condition ( 1 ) in the definition of filtered tableaux .
@xmath88{figures / t1.1.eps}\ \
\ \includegraphics[height=50pt]{figures / t1.2.eps}\ \
\ \includegraphics[height=50pt]{figures / t1.3.eps}\ \ \raisebox{41pt}{$\subset$}\
\ \includegraphics[height=50pt]{figures / t1.4.eps}\
\ \raisebox{41pt}{$\subset$}\ \ \includegraphics[height=50pt]{figures / t1.eps}\ ] ] @xmath88{figures / t1.1.eps}\ \ \raisebox{41pt}{$\subset$}\
\ \includegraphics[height=50pt]{figures / t2.2.eps}\ \ \raisebox{41pt}{$\subset$}\
\ \includegraphics[height=50pt]{figures / t2.3.eps}\ \ \raisebox{41pt}{$\subset$}\
\ \includegraphics[height=50pt]{figures / t2.4.eps}\ \ \raisebox{41pt}{$\subset$}\
\ \includegraphics[height=50pt]{figures / t2.eps}\ ] ] @xmath88{figures / t1.1.eps}\ \ \raisebox{41pt}{$\subset$}\
\ \includegraphics[height=50pt]{figures / t2.2.eps}\ \ \raisebox{41pt}{$\subset$}\
\ \includegraphics[height=50pt]{figures / t3.3.eps}\ \ \raisebox{41pt}{$\subset$}\
\ \includegraphics[height=50pt]{figures / t3.4.eps}\ \
\raisebox{41pt}{$\subset$}\ \
\includegraphics[height=50pt]{figures / t3.eps } \raisebox{-5pt}{\rule{0pt}{0pt}}\ ] ] each of the first two sequences support a unique filtered tableau satisfying condition ( 2 ) , while the third supports two ; thus the required intersection number is 4 , which may be verified by direct computation using the pieri formula for flag manifolds @xcite .
indeed , there is a unique littlewood - richardson tableau of shape @xmath47 and content @xmath28 when @xmath28 is a single row or column and also when the shapes of @xmath47 and @xmath28 are the same or rotated by @xmath89 .
the only skew shape here which admits more than one littlewood - richardson tableau is when @xmath90{figures/21.3.eps}$ ] and @xmath91{figures/1.1.1.eps}$ ] .
there are two such littlewood - richardson tableaux , given in , and this occurs in the middle of the third chain .
let @xmath92}$ ] be the manifold of complete flags in @xmath6 , which has dimension @xmath93 .
its schubert classes are indexed by all permutations @xmath12 of the numbers @xmath94 .
we prove a strengthening of theorem [ thm : main ] for the full flag manifold and use this to deduce theorem [ thm : main ] for all partial flag manifolds .
we give the key definition of this section .
a permutation @xmath12 is a with if @xmath95{figures / qed.eps}}}}\ ] ] for example , @xmath96 and @xmath97 are valley permutations with floor at 3 .
we associate a shape @xmath98 to any valley permutation @xmath12 . if @xmath12 has floor at @xmath99 , then @xmath100 is the shape whose rows are @xmath101 this has either @xmath99 or @xmath102 rows .
observe that @xmath12 is determined by @xmath100 and that @xmath103 where @xmath16 counts the inversions in @xmath12 .
for example , @xmath104{figures / shift42.eps } \qquad\raisebox{19pt}{and}\qquad \raisebox{19pt}{$\mu(64\red{3}125)\ = \ $ } \
\includegraphics[height=30pt]{figures / shift532.eps}\ \raisebox{19pt}{.}\ ] ] [ th : coefficients ] let @xmath105 with @xmath106 and suppose that @xmath12 is a valley permutation with shape @xmath107 . then the coefficient of @xmath14 in the product @xmath108 in the cohomology ring of @xmath109 is the number of filtered tableau with shape @xmath107 and content @xmath83 . since the class @xmath63}$ ] of a point in @xmath110
is indexed by the longest permutation , which is a valley permutation with shape @xmath30{figures / shift.eps}}}_{[n{-}1]}$ ] , theorem [ th : coefficients ] implies theorem [ thm : main ] for @xmath111}$ ] .
we deduce theorem [ thm : main ] for general flag manifolds @xmath11 from the case for @xmath111}$ ] .
suppose that @xmath112 , say @xmath113 , and set @xmath114 .
we assume that the theorem holds for @xmath115 , and deduce it for @xmath116 .
let @xmath117 be the rectangular partition with @xmath118 rows and @xmath119 columns .
set @xmath120 .
note that @xmath121 is dual to @xmath122}_\alpha$ ] in @xmath123 under the poincar pairing .
thus , for any @xmath124 we have @xmath125}_{\alpha'}\big]\ , \pi_{\alpha',b}^ * \sigma_\kappa \cdot \pi_{\alpha',\alpha}^ * \tau \ = \
\big[{[\mathrm{pt}]}_\alpha\big]\ , \tau\,,\ ] ] where @xmath126}_\alpha\big ] \tau$ ] denotes the coefficient of @xmath63}_\alpha$ ] in @xmath127 .
in particular , @xmath128}_{\alpha'}\big ] \prod_{(a , \lambda ) \in \lambda } \pi_a^ * \sigma_\lambda \ = \
\big[{[\mathrm{pt}]}_\alpha\big ] \prod_{(a ' , \lambda ' ) \in \lambda ' } \pi_{a'}^ * \sigma_{\lambda'}\,.\ ] ] there is a bijection between filtered tableaux with shape @xmath30{figures / shift.eps}}}_\alpha$ ] and content @xmath83 and those with shape @xmath30{figures / shift.eps}}}_{\alpha'}$ ] and content @xmath129 , obtained by inserting the unique littlewood - richardson tableau of shape and content @xmath117 into the filtration . thus counting either set of filtered tableaux gives the coefficient .
a schubert class @xmath14 in a product @xmath130 of schubert classes if , when we expand the product in the basis of schubert classes , @xmath14 appears with a positive coefficient . we will prove theorem [ th : coefficients ] by induction on the number of terms @xmath131 in the product .
important for this is the following proposition which summarizes some discussion at the beginning of section 1 in @xcite .
[ p : a - bruhat ] if a schubert class @xmath14 appears in the product @xmath132 , then the following conditions hold .
* whenever @xmath133 , we have @xmath134 and @xmath135 . * if @xmath136 and @xmath137 , then @xmath138 . if @xmath139 and @xmath137 , then @xmath138 . in @xcite
, it is shown that the conditions in proposition [ p : a - bruhat ] define an order relation @xmath140 , which is a suborder of the bruhat order .
we deduce an important lemma .
[ lem : descents ] if @xmath14 appears in @xmath108 then @xmath12 has no descents after position @xmath141 .
we prove this by induction on @xmath131 .
it holds when @xmath142 , as the multiplicative identity in cohomology is the schubert class indexed by the identity permutation .
suppose that @xmath14 appears in the product @xmath108 .
then there is some permutation @xmath143 such that @xmath144 appears in the product @xmath145 and @xmath14 appears in the product @xmath146 .
hence @xmath147 .
since @xmath143 has no descents after position @xmath148 and @xmath149 , condition ( 2 ) of proposition [ p : a - bruhat ] implies that @xmath12 has no descents after position @xmath141 . for permutations @xmath150 and a partition @xmath54{figures / rect.eps}}}_a$ ] ,
let @xmath151 be the coefficient of @xmath14 in the product @xmath152 .
one of the main results in @xcite is the following identity .
[ p : identity ] suppose that @xmath140 and @xmath153 with @xmath154 .
then for every @xmath155{figures / rect.eps}}}_a$ ] we have @xmath156 .
suppose that a shape @xmath157{figures / shift.eps}}}_{[n-1]}$ ] has either @xmath158 or @xmath23 rows .
we define @xmath159 to be the intersection of the shape
@xmath160 with @xmath30{figures / rect.eps}}}_b$ ] .
we proceed by induction on @xmath131 .
the theorem holds ( trivially ) for @xmath142 ; assume that @xmath161 and that it holds for @xmath162 .
let @xmath12 be a valley permutation with shape @xmath107 , and suppose that @xmath12 appears in the product @xmath163 .
then by lemma [ lem : descents ] , @xmath12 has a floor at @xmath141 .
let us expand the product @xmath164 then the coefficient of @xmath14 in the product @xmath163 is the sum @xmath165 suppose that @xmath147 .
since @xmath12 has a floor at @xmath141 , proposition [ p : a - bruhat](2 ) implies that @xmath166 if the coefficient @xmath167 , so that @xmath143 can contribute to this sum , then lemma [ lem : descents ] implies that @xmath143 has no descents after position @xmath148 . since @xmath168 , this implies that @xmath143 is a valley permutation with a floor at @xmath141 .
let @xmath160 be the shape of @xmath143 .
since both @xmath12 and @xmath143 have floor at @xmath141 , both @xmath107 and @xmath160 have either @xmath169 or @xmath141 rows , and thus @xmath170{figures / rect.eps}}}_{a_t}$ ] .
the theorem would follow if we knew that @xmath171 to see this , note that there is a bijection between filtered tableaux on @xmath107 with content @xmath172 and triples @xmath173 where @xmath174 , @xmath175 is a filtered tableau of shape @xmath160 and content @xmath176 , and @xmath177 is a littlewood - richardson tableau of shape @xmath178 and content @xmath28 ; hence the number of these is @xmath179 but follows from proposition [ p : identity ] .
let @xmath180 ( respectively @xmath181 ) be the permutation obtained from @xmath143 ( respectively from @xmath12 ) by reversing the first @xmath141 values , i.e. @xmath182 then @xmath180 and @xmath181 are grassmannian permutations with descent @xmath141 , and shapes @xmath183 and @xmath184 , respectively , and @xmath185 . furthermore , @xmath186 and @xmath154 , from which we deduce .
when all the classes @xmath187 have degree 2 ( @xmath188{figures / box.eps}}}$ ] , a single box ) , the multiplication formula @xmath189{figures / box.eps}}}$ ] is due to monk @xcite .
monk s formula states that @xmath190{figures / box.eps}}}\ = \sum_{\substack{j \leq i < k \\
\ell(w r_{jk } ) = \ell(w)+1 } } \sigma_{w r_{jk}}\,,\ ] ] where @xmath191 is the transposition swapping @xmath48 and @xmath192 .
iterating monk s formula one sees that the coefficient of @xmath63}_\alpha$ ] in a product @xmath193{figures / box.eps}}}$ ] is obtained by counting certain chains in the bruhat order .
it is not hard to see directly from that each permutation @xmath12 in such a chain corresponds to a shape @xmath107 in @xmath30{figures / shift.eps}}}_\alpha$ ] such that the number of boxes in the column @xmath48 of @xmath107 equals @xmath194\ |\ w(k)>w(j+1)\}$ ] , for all @xmath195 . indeed ,
if the permutation @xmath12 does not correspond to a shape , then no term on the right hand side of corresponds to a shape .
it follows that the coefficient is the number of chains of shapes in @xmath30{figures / shift.eps}}}_\alpha$ ] where the @xmath33th step involves adding a box in @xmath30{figures / rect.eps}}}_{a_i}$ ] , which is the answer given by our formula . for example , we have @xmath196}_{[3]}\ = \ \pi^*_1 \includegraphics{figures / b1.eps}\cdot \pi^*_1 \includegraphics{figures / b2.eps}\cdot \pi^*_2 \includegraphics{figures / b3.eps}\cdot \pi^*_2 \includegraphics{figures / b4.eps}\cdot \pi^*_3 \includegraphics{figures / b5.eps}\cdot \pi^*_3 \includegraphics{figures / b6.eps}\ , \ ] ] as there are two chains of shapes which satisfy this condition .
@xmath197{figures / m1.eps}\ \ \raisebox{21pt}{$\subset$}\
\ \includegraphics[height=30pt]{figures / m2.eps}\ \ \raisebox{21pt}{$\subset$}\
\ \includegraphics[height=30pt]{figures / m3.1.eps}\ \ \raisebox{21pt}{$\subset$}\
\ \includegraphics[height=30pt]{figures / m4.1.eps}\ \ \raisebox{21pt}{$\subset$}\
\ \includegraphics[height=30pt]{figures / m5.1.eps}\ \ \raisebox{21pt}{$\subset$}\
\ \includegraphics[height=30pt]{figures / m6.1.eps}\ ] ] @xmath197{figures / m1.eps}\ \ \raisebox{21pt}{$\subset$}\
\ \includegraphics[height=30pt]{figures / m2.eps}\ \ \raisebox{21pt}{$\subset$}\
\ \includegraphics[height=30pt]{figures / m3.2.eps}\ \ \raisebox{21pt}{$\subset$}\
\ \includegraphics[height=30pt]{figures / m4.2.eps}\ \ \raisebox{21pt}{$\subset$}\
\ \includegraphics[height=30pt]{figures / m5.2.eps}\ \ \raisebox{21pt}{$\subset$}\
\ \includegraphics[height=30pt]{figures / m6.2.eps}\smallskip\ ] ] it is possible to give a purely combinatorial proof of theorem [ thm : main ] using kogan s formula ( * ? ? ? * theorem 2.4 ) .
this rule is based on insertion of rc - graphs and gives the coefficient @xmath151 , when @xmath198 .
in particular , this gives a formula for the product when @xmath143 and @xmath12 are a valley permutations with a floor at @xmath99 , and so we may use this in a formula for the intersection numbers of theorem [ thm : main ] to give a combinatorial proof .
the conventions in @xcite for schubert classes differ from those used in this article . to compare conventions ,
it is necessary to replace our permutations @xmath12 by @xmath199 throughout . in particular ,
a cohomology class indexed by @xmath12 in this article is the class indexed by @xmath200 in @xcite .
thus our condition on @xmath143 becomes @xmath201 , which is the condition found in @xcite . to deduce theorem [ thm : main ] from this formula , we would need to show that , for valley permutations @xmath202 with floor at @xmath99 , kogan s rule for @xmath151 coincides with the littlewood - richardson rule for @xmath203 , where @xmath204 and @xmath205 .
here , @xmath206 is the shape of @xmath143 and @xmath100 is the shape of @xmath12 .
while this is certainly possible , we chose not to pursue this .
consider the following product in @xmath207 , @xmath208{appendix/2.1.eps}})\cdot \pi_2^*(\sigma_{\includegraphics[height=4pt]{appendix/2.2.eps}})\cdot \pi_3^*(\sigma_{\raisebox{-4pt}{\includegraphics[height=8pt]{appendix/22.3.eps}}})\cdot \pi_3^*(\sigma_{\raisebox{-4pt}{\includegraphics[height=8pt]{appendix/21.4.eps}}})\cdot \pi_5^*(\sigma_{\includegraphics[height=4pt]{appendix/1.5.eps}})\cdot \pi_5^*(\sigma_{\raisebox{-8pt}{\includegraphics[height=12pt]{appendix/111.6.eps}}})\cdot \pi_5^*(\sigma_{\raisebox{-8pt}{\includegraphics[height=12pt]{appendix/111.7.eps}}})\,.\ ] ] by theorem [ thm : main ] , the coefficient of @xmath63}$ ] is the number of filtered tableau with content @xmath209{appendix/2.1.eps } ) , ( 2,\includegraphics[height=6pt]{appendix/2.2.eps } ) , ( 3,\raisebox{-2pt}{\includegraphics[height=12pt]{appendix/22.3.eps } } ) , ( 3,\raisebox{-2pt}{\includegraphics[height=12pt]{appendix/21.4.eps } } ) , ( 5,\includegraphics[height=6pt]{appendix/1.5.eps } ) , ( 5,\raisebox{-6pt}{\includegraphics[height=18pt]{appendix/111.6.eps } } ) , ( 5,\raisebox{-6pt}{\includegraphics[height=18pt]{appendix/111.7.eps } } ) ) , $ ] which is 18 : @xmath210{appendix / t1.eps } } \end{picture } \qquad \begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t2.eps } } \end{picture } \qquad \begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t3.eps } } \end{picture } \qquad \begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t4.eps } } \end{picture } \qquad \begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t9.eps } } \end{picture}\ ] ] @xmath210{appendix / t10.eps } } \end{picture } \qquad \begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t13.eps } } \end{picture } \qquad \begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t14.eps } } \end{picture } \qquad \begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t15.eps } } \end{picture } \qquad \begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t16.eps } } \end{picture}\ ] ] @xmath210{appendix / t5.eps } } \put(31.5,31){2 } \put(45.5,44.5){1 } \put(58.5,58){1 } \end{picture } \qquad \begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t5.eps } } \put(32,31){1 } \put(45,44.5){2 } \put(58.5,58){1 } \end{picture } \qquad \begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t6.eps } } \put(31.5,31){2 } \put(45.5,44.5){1 } \put(58.5,58){1 } \end{picture } \qquad \begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t6.eps } } \put(32,31){1 } \put(45,44.5){2 } \put(58.5,58){1 } \end{picture}\ ] ] @xmath210{appendix / t7.eps } } \put(31.5,31){2 } \put(45.5,44.5){1 } \put(58.5,58){1 } \end{picture } \qquad \begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t7.eps } } \put(32,31){1 } \put(45,44.5){2 } \put(58.5,58){1 } \end{picture } \qquad \begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t8.eps } } \put(31.5,31){2 } \put(45.5,44.5){1 } \put(58.5,58){1 } \end{picture } \qquad
\begin{picture}(70,70 ) \put(0,0){\includegraphics[height=70pt]{appendix / t8.eps } } \put(32,31){1 } \put(45,44.5){2 } \put(58.5,58){1 } \end{picture}\ ] ] we remarked in section [ sec : comments ] that , when every partition is a single box ( @xmath188{figures / box.eps}}}$ ] ) , a filtered tableau is a particular saturated chain of shapes in @xmath30{figures / shift.eps}}}_\alpha$ ] . when @xmath69 we look at this for the problem @xmath211 in @xmath212 . | we give a combinatorial rule for computing intersection numbers on a flag manifold which come from products of schubert classes pulled back from grassmannian projections .
this rule generalizes the known rule for grassmannians . |
constraint satisfaction problems ( csps ) are computational problems that appear in almost every area of computer science such as artificial intelligence , graph algorithms , scheduling , combinatorics , and computer algebra . depending on the type of constraints that are allowed in the input instances of a csp ,
the computational complexity of a csp is usually polynomial ( we will call these csps tractable ) , or np - hard . in the last decade , a lot of progress was made to find general criteria that imply that a csp is tractable , or that it is np - hard .
such results have been obtained for constraint languages over finite domains @xcite , but also for constraint languages over infinite domains that are @xmath5-categorical ( for formal definition of these concepts see section [ sect : csp ] ) .
for example , it has been shown that for every structure @xmath1 with a first - order definition in @xmath6 the problem csp@xmath7 is in p if it falls into one out of nine classes , and is np - hard otherwise @xcite . lately
, many researchers have been fascinated by a conjecture due to feder and vardi @xcite which is known as the _ dichotomy conjecture_. this conjecture says that every csp with a finite domain constraint language is either tractable ( i.e. , in p ) or np - complete .
according to a well - known result by ladner @xcite , there are np-_intermediate _ computational problems , i.e. , problems in np that are neither tractable nor np - complete ( unless p = np ) .
but the problems that are given in ladner s construction are extremely artificial .
the question why there are so few candidates for natural np - intermediate problems is one of the mysteries in complexity theory .
any outcome of the dichotomy conjecture is probably surprising : a negative answer would finally provide relatively natural np - intermediate problems , which would be of interest in complexity theory .
a positive answer probably comes with a criterion which describes the np - hard csps ( and it would probably even provide algorithms for the tractable csps ) .
but then we would have a rich catalogue of computational problems where the computational complexity is known .
such a catalogue would be a valuable tool for deciding the complexity of computational problems in the mentioned application areas : since csps are abundant , one might derive algorithmic results by reducing the problem of interest to a known tractable csp , and one might derive hardness results by reducing a known np - hard csp to the problem of interest . in this article , we study two natural classes of infinite domain constraint languages , and show that the corresponding csps do exhibit a complexity dichotomy . to the best of our knowledge ,
this is the first systematic complexity result for classes of structures that are not @xmath5-categorical .
the first class consists of all first - order expansions of @xmath8 ( i.e. , we add relations to @xmath8 that are first - order definable in @xmath8 .
the second class is an affine version of the first class , and consists of all first - order expansions of @xmath9 . that the structures @xmath8 and @xmath9 are not @xmath5-categorical follows immediately from the theorem by engeler , ryll - nardzewski , and svenonius ( cf .
theorem 6.3.1 in @xcite ) .
it is even the case that the corresponding csps can not be formulated by any @xmath5-categorical template ; the basic proof idea is presented in ( * ? ? ?
* proposition 1 ) ; also see @xcite .
our results follow from theorems about primitive positive definability : we show that for every relation @xmath10 with a first - order definition in @xmath11 , either @xmath10 has a quantifier - free horn definition in @xmath11 , or @xmath10 contains the tuple @xmath12 , or all relations with a first - order definition in @xmath11 have a primitive positive definition in @xmath13 .
the analogous result also holds for the affine case .
the techniques that we use to prove these two definability theorems are more general than the two classification results , and they are very different in nature .
one technique applies for structures ` that have little structure ' ; to be precise , for _ all _ structures @xmath1 where @xmath14 and @xmath15 are the only primitive positive definable non - trivial binary relations ( section [ sect : affine ] ) .
in particular , they apply to structures with a 2-transitive automorphism group .
the other technique applies for structures ` with a lot of structure ' ; informally , it applies whenever we can find a primitive positive definition for the line between two points in @xmath16 ( section [ sect : q ] ) .
the rest of this paper is organised as follows : in section [ sect : csp ] , we provide some background material on constraint satisfaction and logic . a tractability result for templates that have a quantifier - free horn definition in @xmath4
is presented in section [ sec : tract ] . the classification result for @xmath4
can be found in section [ sect : q ] while the results for the affine case are collected in section [ sect : affine ] . finally , a number of open questions and directions for future work can be found in section [ sect : conc ] .
let @xmath17 be a relational structure with domain @xmath18 ( which will usually be infinite ) and finitely many relations @xmath19 .
the _ constraint satisfaction problem for @xmath1 _ ( short , csp@xmath7 ) is the computational problem to decide whether a given primitive positive sentence @xmath20 involving relation symbols for the relations in @xmath1 is true in @xmath1 .
a first - order formula is called _ primitive positive _ if it is of the form @xmath21 where @xmath22 are atomic formulas , i.e. , formulas of the form @xmath23 or @xmath24 with @xmath10 the relation symbol for a @xmath25-ary relation from @xmath1 .
we call such a formula a _ pp - formula_. the conjuncts in a pp - formula @xmath20 are also called the _ constraints _ of @xmath20 .
we also refer to @xmath1 as a _ constraint language _ ( it is also often called the _ template _ ) of csp@xmath7 .
we say that a first - formula @xmath26 _ defines _ a relation @xmath10 in @xmath1 when @xmath27 holds in @xmath1 iff @xmath28 . if @xmath26 is primitive positive , we call @xmath10 _ primitive positive definable ( pp - definable ) _ over @xmath1
. the following simple but important result explains the importance of primitive positive definability for constraint satisfaction problems .
[ pp - def ] let @xmath1 be a relational structure and @xmath29 be an expansion of this structure by a pp - definable relation @xmath10 over @xmath1 .
then csp@xmath7 is polynomial - time equivalent to csp@xmath30 .
lemma [ pp - def ] will be used extensively in the sequel and we will not make explicit references to it . another important class of formulas are _ horn _ formulas ; a first - order formula in conjunctive normal form is horn if and only if each clause contains at most one positive literal
. a relation @xmath10 is called quantifier - free horn definable over @xmath1 if there exists a quantifier - free horn formula that defines @xmath10 in @xmath1 .
note that lemma [ pp - def ] does not hold if we replace ` pp - definable ' with ` horn definable ' . by choosing an appropriate structure
@xmath1 many computational problems that have been studied in the literature can be formulated as csp@xmath7 ( see e.g. @xcite ) .
it turns out very often that the structure @xmath1 can be chosen to be @xmath5-categorical .
a structure is called _ @xmath5-categorical _ if the set of all first - order sentences that is true in the structure has only one countable model , up to isomorphism .
a famous example of an @xmath5-categorical structure is @xmath31 .
the condition of @xmath5-categoricity is interesting for constraint satisfaction , because the so - called _ universal - algebraic approach _ , which is currently intensively studied for finite constraint languages , applies at least in principle also for @xmath5-categorical structures ( see e.g. @xcite for an application of the universal - algebraic approach to csps for constraint languages over infinite domains ) . in this article , we demonstrate that systematic complexity classification can be performed for constraint languages over infinite domains even if the constraint languages are not @xmath5-categorical .
[ [ example . ] ] * example . *
+ + + + + + + + + + let @xmath1 denote the structure @xmath32 it can be shown that csp@xmath7 _ can not _ be formulated with an @xmath5-categorical template ( for a very similar proof , see @xcite ; a necessary and sufficient condition about which csps can be formulated with @xmath5-categorical templates can be found in @xcite ) .
one can show that the relations @xmath33 and @xmath34 have pp - definitions in @xmath1 .
it is now straightforward to determine the computational complexity of csp@xmath7 by combining lemma [ pp - def ] and our classification result ( corollary [ cor : q ] ) .
we will sometimes consider the automorphism group aut@xmath7 of a template @xmath1 over a domain @xmath18 , i.e. , the group formed by the set of all automorphisms and @xmath1 .
] of @xmath1 with respect to functional composition .
an _ orbit _ of @xmath35 on @xmath36 is a set of the form @xmath37 , for some @xmath38 .
we note that pairs from the same orbit satisfy the same first - order formulas .
let @xmath18 be an arbitrary infinite set and arbitrarily choose an element @xmath39 . the complexity of csp@xmath7 where @xmath1 has a first - order definition in @xmath40 ( so - called _ equality languages _ ) has been classified in @xcite .
we note that if @xmath10 is first - order definable in @xmath40 and @xmath41 , then @xmath42 for every @xmath43 .
thus , the exact choice of @xmath2 is irrelevant when stating the following theorem .
[ thm : ecsp ] let @xmath1 be a template with a first - order definition in @xmath40 . then , all relations in @xmath1 have a quantifier - free horn definition in @xmath40 , or all non - empty relations in @xmath1 contain the tuple @xmath3 , or else every first - order definable relation in @xmath40 has a pp - definition in @xmath1 . in the last case , csp@xmath7 is np - complete . instead of using theorem [ thm : ecsp ] in its full generality , it will be sufficient to use a simple corollary .
for any set @xmath18 , the relation @xmath44 denotes the relation @xmath45 [ s - prop ] let @xmath18 be an infinite set .
every first - order definable relation in @xmath46 has a pp - definition in @xmath47 .
the relation @xmath44 has a first - order definition in @xmath40 and does not contain the tuple @xmath48 .
it is easy to verify that @xmath44 has no quantifier - free horn definition in @xmath40 so every first - order definable relation in @xmath40 has a pp - definition in @xmath47 by theorem [ thm : ecsp ] .
for all relational structures @xmath1 with a quantifier - free horn definition in @xmath4 , the problem csp@xmath7 can be solved in polynomial time .
this follows from a more general algorithmic result in @xcite .
however , the algorithm presented there solves a linear number of linear programs , and thus the best known algorithms have a rather high worst - case running time .
we present a more efficient algorithm for the special case that is relevant in our paper .
we denote by @xmath49 the class of all functions of asymptotic growth at most @xmath50 up to poly - logarithmic factors .
[ prop : tract ] let @xmath1 be a relational structure whose relations have a quantifier - free horn definition in @xmath4 . then there is an algorithm that solves csp@xmath7 in time @xmath51 where @xmath52 is the size of the input .
the algorithm we present in the proof of proposition [ prop : tract ] is a combination of general techniques in constraint satisfaction @xcite and a polynomial implementation of gaussian elimination algorithm on rational data .
since the input of csp@xmath7 consists of a primitive positive sentence whose atomic formulas are of the form @xmath53 where @xmath10 is quantifier - free horn definable over @xmath4 , we can as well assume that the input to our problem consists of a set of horn clauses over @xmath4 .
we have to make some remarks about the worst - case running time of the gaussian elimination algorithm .
it is well - known that the gaussian elimination requires @xmath54 many arithmetic operations on rational numbers , where @xmath55 is the number of equations and @xmath56 is the number of variables . in our algorithm , we have to solve a linear number of linear equation systems @xmath57 ; however , system @xmath58 is obtained from system @xmath59 by adding a single linear equation . since the gaussian algorithm can be presented in such a way that it computes a system in triangular form , adding successively equation by equation , the overall costs for solving @xmath57 equals the cost to solve @xmath60 with gaussian elimination
. also recall that the size of the numbers involved when performing the gaussian elimination algorithm might grow exponentially when implemented without care .
however , when we use the euclidean algorithm to shorten the coefficients during the elimination process , the gaussian elimination algorithm can be shown to be polynomial @xcite .
we are only interested in deciding solvability of linear equation systems , and not constructing solutions , and so we even have _ linear _ bounds ( in the input size ) on the representation size of all numbers involved in deciding solvability for linear equation systems over the rational numbers with gaussian elimination ( see @xcite , proof of theorem 3.3 ) .
finally we remark that the most costly arithmetic operation that has to be performed on rational numbers during the elimination process is multiplication , and multiplication can be performed in time @xmath61 , where @xmath62 denotes the representation size of the two rational numbers ( in bits ) .
hence , the overal running time for solving @xmath57 with the discussed implementation of the gaussian elimination algorithm is in @xmath51 .
we will show that our algorithm for csp@xmath7 can be implemented such that it has the same overall asymptotic worst - case complexity .
we first discuss the correctness of the algorithm shown in figure [ fig : alg ] , and then explain how to implement the algorithm such that it achieves the desired running time .
when @xmath63 logically implies @xmath26 then the negative literal @xmath64 is never satisfied and can be deleted from all clauses without affecting the set of solutions . since this is the only way how literals can be deleted from clauses , it is clear that if one clause becomes empty the instance is unsatisfiable .
if the algorithm terminates with _ yes _
, then no negation of a disequality is implied by @xmath63 .
if @xmath65 is the rank of the linear equation system defined by @xmath63 , we can use the gaussian elimination algorithm as described above to eliminate from all literals in the remaining clauses @xmath65 of the variables .
let @xmath66 be the maximal sum of the absolute values of all coefficients in one of the remaining inequalities plus one .
then setting the @xmath67-th variable to @xmath68 satisfies all clauses . to see this ,
take any disequality , and assume that @xmath67 is the highest variable index in this disequality .
order the disequality in such a way that the variable with highest index is on one side and all other on the other side of the @xmath15 sign .
the absolute value on the side with the @xmath67-th variable is at least @xmath68 .
the absolute value on the other side is less than @xmath69 , since all variables have absolute value less than @xmath70 and the sum of all coefficients is less than @xmath71 in absolute value .
hence , both sides of the disequality have different absolute value , and the disequality is satisfied . since all remaining clauses have at least one disequality , all constraints are satisfied .
we finally explain how to implement the algorithm such that it runs in time @xmath51 . to decide whether @xmath63 implies an equality @xmath26 , we compute in each interation of the main loop the triangular normal form for the linear equation system determined by @xmath63 as described before the statement of the proposition . the overall costs to do this are in @xmath51 . moreover ,
for each negative literal we maintain an equation where we eliminate as many variables as possible using the computed triangular normal form .
if one of these equations becomes trivial ( i.e. is the form @xmath72 ) we conclude that the equation is implied by @xmath63 . the overall costs for doing this
is also bounded by @xmath51 by a very similar argument as given before the statement of the proposition . with appropriate straightforward data structures , the total costs for removing negated literals @xmath64 from all clauses when @xmath26 is implied by @xmath63
is linearly bounded in the input size since each literal can be removed at most once .
in this section we present the complexity classification for first - order expansions of @xmath73 . we begin in section [ defneq ] with a result about the pp - definability of the disequality relation @xmath15 in first - order expansions of @xmath73 . when the relation @xmath15 is pp - definable ,
we show that also the relation @xmath74 ( defined in section [ sect : csp ] as the relation @xmath75 ) is pp - definable whenever the constraint language contains a relation @xmath10 that is first - order , but not quantifier - free horn definable in @xmath11 ; this is shown in section [ sect : defsq ] .
finally , section [ sect : cc1 ] completes the classification for first - order expansions of @xmath73 .
[ lem : trans ] for any structure @xmath1 with a first - order definition in @xmath76 , the first - order definable relations in @xmath1 are a subset of @xmath77 .
let @xmath10 be a unary relation with a first - order definition in @xmath76 .
the statement is clear if @xmath10 does not contain any element distinct from @xmath78 , so let @xmath79 be from @xmath80 .
we have to show that @xmath81 or @xmath82 .
observe that for any @xmath83 , @xmath84 , the mapping @xmath85 is an automorphism of @xmath1 .
hence , for any @xmath86 there is an automorphism of @xmath76 that maps @xmath79 to @xmath87 . since automorphisms preserve first - order formulas ,
so @xmath88 and the claim follows .
note that @xmath89 is equivalent to @xmath90 and hence the relation @xmath91 is pp - definable over @xmath76 ; thus we can use @xmath78 freely as a constant symbol in pp - definitions over @xmath1 .
[ prop : neq ] let @xmath1 be a first - order expansion of @xmath92 containing a non - empty relation @xmath10 such that @xmath93 is false for any @xmath94
. then @xmath15 is pp - definable in @xmath1 .
observe that if the set @xmath80 has a pp - definition @xmath95 in @xmath1 , then the pp - formula @xmath96 defines @xmath97 over @xmath1 .
let @xmath66 be a non - empty pp - definable relation in @xmath1 of minimal arity such that @xmath98 defines the empty set .
let @xmath25 be the arity of @xmath66 .
first , assume that @xmath99 is satisfiable .
then the @xmath100-ary relation @xmath101 defined by @xmath102 is non - empty , and @xmath103 defines the empty set ; this is in contradiction to the choice of @xmath66 .
assume next that @xmath104 is unsatisfiable .
define the unary relation @xmath105 by @xmath106 and the unary relation @xmath107 by @xmath108 by lemma [ lem : trans ] , both @xmath109 and @xmath110 are from @xmath77 .
the relation @xmath109 can not be equal to @xmath91 or to @xmath111 since this contradicts the assumption that @xmath99 is unsatisfiable . if @xmath109 is equal to @xmath80 , then by the initial observation @xmath15 is pp - definable in @xmath1 and we are done .
we conclude that @xmath112 and hence @xmath113 .
since @xmath110 is non - empty , it must be the case that @xmath114 , and again by the initial observation @xmath15 is pp - definable in @xmath1 .
the rational numbers with addition ( and also the real numbers with addition ) admit _ quantifier elimination _ , i.e. , every relation with a first - order definition in @xmath11 also has a quantifier - free definition over @xmath11 .
this follows from the more general fact that the first - order theory of torsion - free divisible abelian groups admits quantifier elimination ( see e.g. theorem 3.1.9 in @xcite ) .
the first lemma allows us to freely use certain expressions in pp - definitions over @xmath76 .
[ lem : pp ] the relation @xmath115 is pp - definable in @xmath76 for arbitrary @xmath116 . first observe that we can assume that @xmath117 are integers , because we can multiply the equation @xmath118 by the least common multiple of the denominators of @xmath119 and obtain an equivalent equation .
the proof is by induction on @xmath120 .
we first consider that case that @xmath121 .
if @xmath122 , there is nothing to show . otherwise , the formula @xmath123 is equivalent to @xmath124 .
hence , we can in particular use expressions of the form @xmath89 and @xmath125 in pp - definitions over @xmath4 with variables @xmath126 .
if @xmath127 , and @xmath122 or @xmath128 , then we can argue as in the case @xmath121 .
if @xmath129 and @xmath130 are both positive or both negative , then @xmath131 is equivalent to @xmath132 if @xmath129 and @xmath130 have different signs , we replace the conjunct @xmath133 in the formula above by @xmath134 .
now suppose that @xmath135 . by the inductive assumption
, there is a pp - definition @xmath136 for @xmath137 and a pp - definition @xmath138 for @xmath139 .
then @xmath140 is a pp - definition for @xmath118 . in the following
, @xmath10 denotes a relation with a quantifier - free first - order definition @xmath26 in @xmath11 .
a quantifier - free first - order formula @xmath26 in conjunctive normal form is called _ reduced _ if every formula obtained from @xmath26 by removing a literal is not equivalent to @xmath26 ( this concept was introduced in @xcite ) .
clearly , such a reduced definition of @xmath10 always exists , because we can find one by successively removing literals from @xmath26 .
note that if @xmath120 is a literal from @xmath26 , then @xmath141 can be written as a pp - formula over a structure that contains @xmath15 and @xmath142 .
[ lem : hard ] if @xmath10 is first - order , but not quantifier - free horn definable in @xmath11 , then @xmath74 has a pp - definition in @xmath143 .
let @xmath144 be the binary relation defined by @xmath145 .
we first prove that @xmath109 has a pp - definition in @xmath143 .
let @xmath26 be a reduced first - order definition of @xmath10 , and let @xmath146 be a clause of @xmath26 with two positive literals @xmath147 and @xmath148 .
because @xmath26 is reduced , there are @xmath149 such that @xmath150 satisfies @xmath147 and does not satisfy all other literals in @xmath146 , and @xmath151 satisfies @xmath148 but does not satisfy all other literals in @xmath146 .
we claim that the following pp - formula is logically equivalent to @xmath145 .
@xmath152 let @xmath153 be arbitrary .
suppose that @xmath154 .
then the assignment @xmath155 obviously satisfies the first line in the pp - formula .
recall that @xmath156 and @xmath150 does not satisfy all literals in @xmath146 except for @xmath147 .
the function @xmath157 is in aut@xmath11 whenever @xmath153 .
consequently , @xmath158 , too , and the second line in the formula is satisfied as well .
now suppose that @xmath23 .
then the assignment @xmath159 obviously satisfies the first line in the pp - formula . by construction , @xmath160 and @xmath151
does not satisfy all literals in @xmath146 except for @xmath147 .
again we conclude that the second line in the formula is also satisfied . for the opposite direction , suppose that @xmath161 satisfy the pp - formula .
because of the first line of the formula , @xmath153 .
let @xmath162 be the @xmath25 elements whose existence is asserted in the first line of the formula .
note that the equations of the first line imply that @xmath163 lies on the line @xmath164 defined by @xmath165 and @xmath166 . because the formula contains the conjunct @xmath167 , the clause @xmath146 in @xmath26 is satisfied by @xmath162 .
since @xmath162 also satisfies the conjunction of all negated literals in @xmath146 except for the positive literals @xmath147 and @xmath148 , at least one of these two literals @xmath147 and @xmath148 must be satisfied by @xmath162 .
suppose first that @xmath147 is satisfied .
the line @xmath168 does not lie completely within the subspace of @xmath16 defined by @xmath147 ( because @xmath151 does not satisfy @xmath147 , and neither does @xmath169 ) .
hence , @xmath168 intersects this subspace in at most one point . because @xmath150 and hence also @xmath170 satisfies @xmath147 ,
we have thus shown that @xmath163 equals @xmath165 . since @xmath171
we conclude that @xmath154 by the equations in the second line of the formula .
now , consider the case that @xmath148 is satisfied .
similarly as in the last case , @xmath168 intersects the subspace defined by @xmath148 in at most one point . because @xmath172 satisfies @xmath148 , we have shown that @xmath163 equals @xmath151 .
the equations in the second line of the formula then imply that @xmath23 .
finally , we prove that @xmath173 has the following pp - definition in @xmath174 : @xmath175 suppose first that @xmath176 .
note that @xmath177 is not equal to @xmath78 because @xmath178 .
if @xmath179 , then @xmath154 , and if @xmath180 , then @xmath181 so @xmath182 is satisfied . conversely , suppose that @xmath183 satisfies the pp - formula above .
the formula @xmath182 implies that @xmath153 and hence @xmath184 .
moreover , @xmath182 implies that @xmath154 or @xmath23 . if @xmath154 , then @xmath179 and @xmath176 .
if @xmath23 , then @xmath185 and hence @xmath180 . again @xmath186 is in @xmath74 .
we will now use lemma [ lem : hard ] in order to prove the following definability result .
[ thm : q ] let @xmath1 be first - order expansion of @xmath92 .
then , either * each relation in @xmath1 has a quantifier - free horn definition in @xmath92 , or * every non - empty relation of @xmath1 contains a tuple of the form @xmath12 , or * every first - order definable relation in @xmath92 has a pp - definition in @xmath1 .
suppose that there is a non - empty @xmath25-ary relation @xmath10 of @xmath1 that does not contain the tuple @xmath12 .
then the @xmath187-ary relation @xmath188 defined by @xmath189 is non - empty , and the relation defined by @xmath190 is empty .
so we can apply proposition [ prop : neq ] and find that @xmath15 is pp - definable in @xmath191 and hence also in @xmath1 .
so assume in the following without loss of generality that @xmath1 contains the relation @xmath15 .
suppose that one of the relations of @xmath1 does not have a quantifier - free horn definition in @xmath92 .
lemma [ lem : hard ] implies that the relation @xmath74 has a pp - definition in @xmath1 , and corollary [ s - prop ] implies that every relation with a first - order definition in @xmath192 has a pp - definition in @xmath1 .
let @xmath10 be a relation with a first - order definition @xmath26 in @xmath92 . to find a pp - definition for @xmath10 in @xmath1
, we introduce a variable @xmath193 for every atomic formula of the form @xmath194 in @xmath26 . for each atomic formula @xmath195 in @xmath26 of the form @xmath194
, we replace @xmath195 by @xmath196 for a new variable @xmath197 .
the resulting formula consists of a boolean combination of atomic formulas of the form @xmath23 , which we know has a pp - definition @xmath198 in @xmath1 . for each atomic formula @xmath195 in @xmath26
we add the conjunct @xmath199 to @xmath198 , and finally existentially quantify over all new variables .
it is straightforward to verify that the resulting formula is a pp - definition of @xmath10 in @xmath1 .
theorem [ thm : q ] has immediate consequences for the computational complexity of constraint satisfaction .
[ cor : q ] let @xmath1 be a structure with a finite relational signature and a first - order definition in @xmath76 that contains the relation @xmath200 . then csp@xmath7 is in p if all relations in @xmath1 have a quantifier - free horn definition over @xmath76 , or if all non - empty relations contain a tuple of the form @xmath12 , and is np - hard otherwise .
if all relations in @xmath1 have a quantifier - free horn definition over @xmath4 , then proposition [ prop : tract ] implies that csp@xmath7 is in p. otherwise , theorem [ thm : q ] implies that in particular the relation defined by @xmath201 is pp - definable in @xmath1 .
it follows from theorem [ thm : ecsp ] that the constraint satisfaction problem for this ternary relation is np - hard .
we will now consider affine additive structures over @xmath111 .
the structure of this section is very similar to the structure of section [ sect : q ] : we begin by studying the definability of @xmath15 ( section [ sect : neq2 ] ) and of @xmath44 ( in section [ sect : defsd ] ) and use these results to completely classify the problem in section [ sec : cc2 ] . the main proof in section [ sect : defsd ] , however , is very different from the corresponding proof in section [ sect : defsq ] .
let us now formally define the problem at hand : define the operation @xmath202 by @xmath203 .
we study the constraint satisfaction problem for templates @xmath1 with a first - order definition in @xmath204 that contain the relation @xmath205 .
[ lem:2trans ] let @xmath1 be a structure with a first - order definition in @xmath204 .
then there are at most four first - order definable binary relations : the empty relation , the full relation , the relation @xmath15 , and the relation @xmath14 .
it suffices to show that aut@xmath7 has precisely two orbits on @xmath206 , namely @xmath207 these two orbits clearly partition @xmath208 .
it is obvious that @xmath209 is an orbit , because for every @xmath83 the mapping @xmath210 is an automorphism of @xmath204 and hence of @xmath1 . to see that @xmath211 is an orbit of pairs of reals ,
we apply linear interpolation : let @xmath212 and @xmath213 be arbitrary .
the mapping @xmath214 maps @xmath215 to @xmath216 and it is an automorphism of @xmath204 , and hence of @xmath1 . in the proof of lemma [ lem:2trans ] we have in fact verified that the automorphism group of @xmath1 is _ 2-transitive _ , i.e. , that there is only one orbit of pairs of distinct elements with respect to the componentwise action of the automorphism group of @xmath1 on pairs .
[ thm : neq2 ] let @xmath1 be a relational structure with a 2-transitive automorphism group . if there is no pp - definition of @xmath15 , then there is an element @xmath94 of @xmath1 such that every non - empty relation in @xmath1 contains a tuple of the form @xmath217 .
the central step of the classification is the following result concerning pp - definability .
[ lem : indep2 ] let @xmath1 be a relational structure over an infinite domain @xmath18 such that @xmath36 , @xmath14 , @xmath15 , and @xmath218 are the only pp - definable binary relations .
suppose that @xmath1 contains a relation @xmath219 such that there are pairwise distinct @xmath220 for which the following conditions hold : 1 .
@xmath221 is satisfiable ; 2 .
@xmath222 is satisfiable ; 3 .
@xmath223 is unsatisfiable .
then @xmath44 has a pp - definition in @xmath1 .
we simplify the proof of lemma [ lem : indep2 ] by first proving a slightly restricted version : [ lem : indep1 ] let @xmath1 be a relational structure over an infinite domain @xmath18 such that @xmath36 , @xmath14 , @xmath15 , and @xmath218 are the only pp - definable binary relations .
suppose that @xmath1 contains a relation @xmath219 such that there are @xmath224 for which the following conditions hold : 1 .
@xmath221 is satisfiable ; 2 .
@xmath225 is satisfiable ; 3 .
@xmath226 is unsatisfiable .
then @xmath44 has a pp - definition in @xmath1 .
the indices @xmath227 must be pairwise distinct , so suppose for the sake of notation that @xmath228 , @xmath229 , @xmath230 .
consider the relation @xmath10 defined by @xmath231 we first note that @xmath10 is a non - empty relation : @xmath232 is satisfiable so the only way of making @xmath10 empty is that every tuple @xmath233 in @xmath219 satisfies @xmath234 .
this is impossible since we know that there exists a tuple @xmath235 such that @xmath236 .
this implies @xmath237 and contradicts the third condition .
arbitrarily choose a domain element @xmath79 .
we first show that there always exist elements @xmath238 such that @xmath239 .
let @xmath240 and note that @xmath241 is pp - definable .
we know that @xmath241 is non - empty since @xmath10 is non - empty .
now assume that @xmath242 .
first suppose that @xmath243 .
then @xmath244 is non - empty and a strict subset of the equality relation , a contradiction .
if @xmath245 , then consider the pp - definable relation @xmath246 we see that @xmath247 which contradicts the fact that the only non - trivial binary relations that are pp - definable from @xmath1 are @xmath14 and @xmath15 . hence , @xmath248 .
we now continue by considering the tuple @xmath239 . by the third condition
, we see that at least one of @xmath249 must equal @xmath79 in order to satisfy @xmath10 .
let us consider the case @xmath250 .
note that @xmath251 due to the literal @xmath252 .
we now show that @xmath250 is satisfied by any choice of @xmath253 except @xmath79 . to see this ,
assume to the contrary that there is a domain element @xmath254 such that @xmath255 .
define @xmath256 and note that @xmath257 which contradicts the assumption that @xmath14 and @xmath258 are the only non - trivial pp - definable binary relations .
similarly , one can show that @xmath259 holds for all @xmath260
. therefore @xmath261 .
assume for notational simplicity that @xmath262 , and @xmath263 .
define the @xmath264-ary relation @xmath10 by @xmath265 and consider the formula @xmath266 we claim that @xmath267 and @xmath268 are satisfiable while @xmath269 is not satisfiable . then we can apply lemma [ lem : indep1 ] and are done .
first we make an observation : define relation @xmath270 such that @xmath271 we know that @xmath272 is satisfiable so @xmath270 is a non - empty relation . since @xmath273 is not satisfiable ,
we conclude that @xmath270 is a non - empty subset of the equality relation .
consequently , @xmath270 is the equality relation .
analogously , define @xmath274 such that @xmath275 and note that @xmath274 is the equality relation , too .
we now prove that @xmath269 is not satisfiable . by using observation 1
, it follows that any solution @xmath62 satisfies @xmath276 and @xmath277 this is impossible due to the clause @xmath278 .
next , we prove that @xmath267 is satisfiable ; the case @xmath268 is symmetric .
consider the relation @xmath279 by the conditions on @xmath10 , we know that @xmath110 is non - empty . since @xmath110 is binary , we also know that @xmath110 either is the equality relation , the disequality relation , or the full relation .
we conclude that @xmath110 is non - empty and symmetric . by observation 1 , the clause @xmath97 has the effect that every solution @xmath62 must satisfy @xmath280 .
the solution also has to satisfy @xmath278 which implies that @xmath281 .
observation 1 now tells us that @xmath282 and we conclude that every solution satisfies @xmath280 and @xmath282 .
we define @xmath283 thus , @xmath198 is satisfiable if and only if @xmath267 is satisfiable .
we will now construct a concrete satisfying assignment @xmath62 to the variables of @xmath198 . arbitrarily choose a tuple @xmath284 and let @xmath285 , @xmath286 . by the conditions on @xmath110
, there exists an element @xmath287 such that @xmath288 and @xmath289 ; we let @xmath290 .
furthermore , we know that @xmath291 and @xmath292 so @xmath293 and @xmath294 . at this point
, we see that the assignment @xmath62 satisfies the clauses @xmath295 and @xmath97 . we know that @xmath284 so @xmath296 , too , and there exists a value @xmath2 such that @xmath297 and @xmath298
now , let @xmath299 and note that @xmath300 is satisfied by @xmath62 . finally , @xmath301 so the clause @xmath278 is satisfied and the proof is completed .
we are now ready to prove the classification result for the affine case .
[ thm : affine ] let @xmath1 be a first - order expansion of @xmath204 .
then , either * each relation in @xmath1 has a quantifier - free horn definition in @xmath204 , or * every non - empty relation of @xmath1 contains a tuple of the form @xmath12 , or * every first - order definable relation in @xmath204 has a pp - definition in @xmath1 .
suppose that there is a non - empty @xmath25-ary relation @xmath10 of @xmath1 that does not contain the tuple @xmath12 .
the proof of lemma [ lem:2trans ] shows that @xmath1 is 2-transitive , and hence by the contraposition of theorem [ thm : neq2 ] the relation @xmath15 is pp - definable .
so assume in the following without loss of generality that @xmath1 contains the relation @xmath15 .
let @xmath10 be a relation in @xmath1 that does not have a quantifier - free horn definition in @xmath302 .
let @xmath303 be a reduced definition of @xmath10 in @xmath302 ( see section [ sect : q ] ) .
then there must be a clause @xmath146 in @xmath26 with at least two positive literals @xmath304 and @xmath305 .
let @xmath306 be the relation defined by @xmath307 this relation @xmath219 is clearly pp - definable over @xmath308 .
we claim that @xmath219 satisfies the conditions of lemma [ lem : indep2 ] ( which is applicable due to lemma [ lem:2trans ] ) with respect to the arguments indexed by @xmath309 , and @xmath310 ( or the conditions of lemma [ lem : indep1 ] if @xmath311 ; this remark also applies to all other places where we appeal to lemma [ lem : indep2 ] ) .
since @xmath26 is reduced , there is a tuple @xmath312 that satisfies @xmath148 and does not satisfy all other literals in @xmath26 .
now , the extended tuple @xmath313,\dots , t[n],t[i_4],t[j_4])$ ] clearly satisfies @xmath219 , and we have @xmath314 \neq t_1[n+2]$ ] as required in the conditions for lemma [ lem : indep1 ] .
there is also a tuple @xmath315 that satisfies @xmath147 and does not satisfy all other literals in @xmath146 , and we can argue similarly to find a second tuple showing the second condition of lemma [ lem : indep2 ] .
finally , suppose for contradiction that there is a tuple @xmath316 in @xmath219 where @xmath317 \neq t_3[n+1]$ ] and @xmath318 \neq t_3[n+2]$ ] .
because this tuple satisfies in particular the clause @xmath146 from @xmath26 , the conjunct @xmath319 implies that either @xmath147 or @xmath148 is satisfied .
but then the equalites @xmath320 and @xmath321 imply that @xmath317 = t_3[n+1]$ ] or @xmath318 = t_3[n+2]$ ] , a contradiction .
hence , lemma [ lem : indep1 ] applies , @xmath322 is pp - definable over @xmath323 and therefore also over @xmath308 and @xmath1 .
the result follows from corollary [ s - prop ] .
the next corollary is a direct consequence of proposition [ prop : tract ] , theorem [ thm : affine ] , and corollary [ s - prop ] .
let @xmath1 be an expansion of @xmath324 by finitely many first - order definable relations .
if each relation in @xmath1 has a quantifier - free horn definition in @xmath325 , or if each non - empty relation contains a tuple of the form @xmath12 , then csp@xmath7 is in p. otherwise , csp@xmath7 is np - hard .
we have presented classification results for certain algebraic constraint satisfaction problems , and the results are to a large extent based on dichotomy results for logical definability .
we feel that the results and ideas presented in this paper can be extended in many different directions .
hence , it seems worthwhile to provide some concrete suggestions for future work . the results and proof techniques in section [ sect : q ] appear to be generalisable to many different templates defined over various structures .
one example is the natural and important class of structures that are definable in _ presburger arithmetics _
@xcite , i.e. , structures that are first - order definable over the integers with addition @xmath326 .
we note that the following can be obtained by slightly modifying corollary [ cor : q ] .
let @xmath1 be a relational structure with a quantifier - free first - order definition in @xmath327 that contains the relation @xmath200 .
then csp@xmath7 is in p if all relations in @xmath1 have a quantifier - free horn definition over @xmath327 , or if all non - empty relations contain a tuple of the form @xmath12 . otherwise , csp@xmath7 is np - hard .
there is an important difference between this result and a full classification result : we have replaced _ first - order definability _ with _ quantifier - free first - order definability _ in the statement of the result , and the reason is that @xmath328 does not admit quantifier elimination .
is there still a complexity dichotomy if we look at the class of csps with an template that is first - order definable in @xmath326 ?
this appears to be a difficult question .
the results presented in section 5 have strong connections with earlier work on the complexity of _ disjunctive _ constraints @xcite .
we say that @xmath15 is _ 1-independent _ with respect to a @xmath329-structure @xmath1 if and only if for every primitive positive @xmath329-formula @xmath26 with free variables @xmath330 the following holds : if @xmath267 and @xmath331 are satisfiable , then so is @xmath332 .
assume that csp@xmath7 is tractable and let @xmath29 denote the set of all relations that can be defined by ( quantifier - free ) conjunctions of disjunctions over @xmath1 containing at most one literal that is not of the form @xmath97 .
the following has been shown in @xcite ; it does not imply our result since it only makes a statement about a constraint language @xmath29 of the form described above .
let @xmath1 and @xmath29 be defined as above , and assume that p @xmath15 np .
then csp@xmath30 is tractable if and only if @xmath15 is 1-independent with respect to @xmath1 .
we have already mentioned that the structures studied in this paper are in general not @xmath5-categorical . however , torsion - free divisible abelian groups such as @xmath76 and all structures first - order definable in such groups are strongly minimal ( see e.g. corollary 3.1.11 in @xcite ) , and hence categorical in all uncountable cardinals .
this is interesting from a constraint satisfaction point of view because of the following preservation theorem .
[ thm : galois ] let @xmath1 be an uncountably categorical structure with a countable relational signature and an uncountable domain .
then a first - order definable relation @xmath10 has a pp - definition in @xmath1 if and only if @xmath10 is preserved by all _
infinitary _ polymorphisms of @xmath1 .
note that this theorem is weaker than the corresponding theorem for @xmath5-categorical structures @xcite , because we have to assume that @xmath10 is first - order definable , and that @xmath10 is not only preserved by the finitary , but also by the infinitary polymorphisms of @xmath1 . since our classification result is purely in terms of primitive positive definability of first - order definable relations , it is an interesting question to describe the polymorphisms that guarantee tractability for structures @xmath1 with a first - order definition in @xmath4 ( theorem [ thm : galois ] shows that such polymorphisms do exist ) .
we want to thank barnaby martin for comments on an earlier version of the paper .
peter jonsson is partially supported by the _ center for industrial information technology _ ( ceniit ) under grant 04.01 and by the _ swedish research council _
( vr ) under grant 2006 - 4532 . | we study techniques for deciding the computational complexity of infinite - domain constraint satisfaction problems . for certain fundamental algebraic structures @xmath0 , we prove definability dichotomy theorems of the following form : for every first - order expansion @xmath1 of @xmath0 , either @xmath1 has a quantifier - free horn definition in @xmath0 , or there is an element @xmath2 of @xmath1 such that all non - empty relations in @xmath1 contain a tuple of the form @xmath3 , or all relations with a first - order definition in @xmath0 have a primitive positive definition in @xmath1 .
the results imply that several families of constraint satisfaction problems exhibit a complexity dichotomy : the problems are in p or np - hard , depending on the choice of the allowed relations . as concrete examples ,
we investigate fundamental algebraic constraint satisfaction problems .
the first class consists of all first - order expansions of @xmath4 .
the second class is the affine variant of the first class . in both cases ,
we obtain full dichotomies by utilising our general methods . |
classical / nonclassical separation is a controversial subject .
there have been debates on the definition of nonclassical correlation of a multipartite system .
it is well - known that the separability paradigm @xcite is widely - accepted to define entanglement as a quantum correlation that can not be generated from scratch using only local operations and classical communication ( locc ) ( see , e.g. , ref . ) .
there are different paradigms to define nonclassical correlation of a multipartite system from operational viewpoints .
@xcite discussed a certain nonlocality about locally nonmeasurable separable states .
ollivier and zurek @xcite introduced a measure called quantum discord defined as a discrepancy of two expressions of a mutual information that should be equivalent to each other in a classical information theory .
a simple classical / nonclassical separation was given by oppenheim _
@xcite which is also widely recognized .
they defined the class of ( properly ) classically correlated states that are the states with a product eigenbasis .
its complement is the class of nonclassically correlated states that are the states without product eigenbasis .
this definition is in accord with their measure called quantum deficit defined as a discrepancy between the information that can be localized by applying closed locc ( clocc ) operations and the total information of the system .
the clocc protocol allows only local unitary operations , attaching ancillas in separable pure states , and operations to send subsystems through a complete dephasing channel .
thus classically correlated states have vanishing quantum deficit .
other measures @xcite were later proposed on the basis of the same definition of classical / nonclassical correlation .
we aim to evaluate measures based on the following separation of classical / nonclassical correlations .
a quantum bipartite system consisting of subsystems a and b is ( properly ) classically correlated if and only if it is described by a density matrix having a biproduct eigenbasis .
a straightforward extension gives the definition : a quantum multipartite system consisting of subsystems @xmath0 is nonclassically correlated if and only if it is described by a density matrix having no @xmath1-product eigenbasis .
the set of classically correlated states is a nonconvex subset of the set of separable states .
thus a convexity can not be a property of a measure of nonclassical correlation .
the natural statement is that a measure @xmath2 of nonclassical correlation should satisfy the following conditions : + ( i ) @xmath3 if a system is described by a density matrix having a product eigenbasis .
+ ( ii ) @xmath2 is invariant under local unitary operations .
+ these conditions are considered to be prerequisite hereafter .
in addition , the additivity property should be satisfied if one needs to compare systems with different dimensions .
there are two ways to define additivity for a measure .
one is defined as follows , which is valid for any measure of bipartite correlation .
[ defadd1 ] let @xmath4 be a measure of correlation between subsystems a and b of a bipartite system ab , where @xmath5 denotes splitting between a and b. then , @xmath4 is called an additive measure if and only if the equality @xmath6 holds .
another definition of additivity may be introduced for a multipartite measure when we focus on scaling .
[ defadd2 ] let us denote a measure of @xmath1-partite nonclassical correlation by @xmath7 where @xmath8 is a density matrix of an @xmath1-partite quantum system .
first , the measure is fully additive if and only if @xmath9 with @xmath10 the density matrix of an @xmath11-partite system and @xmath12 the density matrix of an @xmath13-partite system .
second , the measure possesses weak additivity if and only if @xmath14 third , the measure possesses subadditivity if and only if @xmath15 in this contribution , we first compare the definitions and properties of measures of nonclassical correlation in sec
. [ sec2 ]
. then we compare these measures numerically in simple examples in sec.[sec3 ] .
the paper concludes with several remarks .
four measures of nonclassical correlation are evaluated for comparison together with the negativity ( an entanglement measure ) @xcite .
the first one is an uncertainty remaining for a third party about the quantum state shared by multiple persons after the third party receives many reports from them @xcite .
these reports are measurement results on the shared state using local observables .
let us consider a density matrix @xmath16}$ ] of an @xmath1-partite system .
consider local complete orthonormal bases @xmath17}\rangle\}_j , \ldots , \{|c_x^{[m]}\rangle\}_x$ ] .
suppose the @xmath18th person locally makes measurements using the observable @xmath19}\rangle\langle c_s^{[l]}|$ ] and sends reports to the third party .
then , a measure of nonclassical correlation is given by @xmath20})= \underset{\mathrm{local~bases}}{\mathrm{min } } \left(-\sum_{j,\ldots , x}p_{j,\ldots , x}\log_2 p_{j,\ldots , x}\right ) - s_\mathrm{vn}(\rho^{[1,\ldots , m]})\ ] ] with @xmath21}|\langle c_k^{[2]}|\cdots \langle c_x^{[m]}| \rho^{[1,\ldots , m]}|c_j^{[1]}\rangle |c_k^{[2]}\rangle\cdots $ ] ( @xmath22 is the von neumann entropy ) .
the value of @xmath23})$ ] vanishes if @xmath16}$ ] has a ( fully ) product eigenbasis .
in addition , @xmath23})$ ] is invariant under local unitary operations as is clear from its definition .
furthermore , it is fully additive in terms of definition [ defadd2 ] .
the second one is derived from a game of multiple persons sharing a quantum state .
let us consider an artificial game to find out eigenvalues of the reduced density matrix of a subsystem from eigenvalues of the density matrix of the total system .
suppose that kate has the @xmath24th component of an @xmath1-partite quantum system .
let the dimension of the hilbert space of the @xmath24th component be @xmath25}$ ] and that of the hilbert space of the total system be @xmath26 .
she wants to know the eigenvalues @xmath27}\}_{j=1}^{d^{[k]}}$ ] of the reduced density matrix of the @xmath24th component .
kate receives @xmath26 eigenvalues from tony who knows all the eigenvalues of the total system .
kate partitions them into @xmath25}$ ] sets .
summing up elements in individual sets , she has @xmath25}$ ] mimic eigenvalues @xmath28}\}$ ] .
thus a measure of nonclassical correlation for kate can be @xmath29})=\underset{{\mathrm{partitionings}}}{{\mathrm{min } } } \left|\sum_{j=1}^{d^{[k]}}(\tilde e_j^{[k]}\log_2 \tilde e_j^{[k ] } - e_j^{[k]}\log_2 e_j^{[k]})\right|.\ ] ] we may take the maximum over @xmath24 to have the measure @xmath30})=\underset{{k}}{{\mathrm{max}}}~f_k(\rho^{[1,\ldots , m]}).\ ] ] this is equal to zero if @xmath16}$ ] has a ( fully ) product eigenbasis .
in addition , it is invariant under local unitary operations as is clear from its definition .
it is subadditive in terms of definition [ defadd2 ] .
the third one is a sort of measures introduced by groisman _
et al_.@xcite .
this is defined in the following way .
consider a bipartite system ( ab ) represented by a density matrix @xmath31 .
then , ( i ) find a basis that diagonalizes the state @xmath32 .
( ii ) write @xmath31 under the basis found by ( i ) and delete all off - diagonal elements .
denote this state as @xmath33 .
( iii ) the measure is calculated by a certain distance between @xmath31 and @xmath33 .
we can take the difference in von neumann entropy as the distance function to define the measure : @xmath34 this measure can be seen as a variant of measure @xmath35 by taking a fixed set of local bases instead of searching for the minimum in ( [ eqdm ] ) .
an extension to the multipartite case is obvious .
this measure satisfies additivity in terms of definition [ defadd1 ] . to define the fourth one
, we can get a clue to construct a measure of nonclassical correlation from a conventional entanglement measure in the following way .
a well known measure of entanglement is negativity @xcite , @xmath36 , defined as the absolute value of the sum of negative eigenvalues of @xmath37 where @xmath38 is the transposition .
this can be in fact regarded as a measure of nonclassical correlation .
nevertheless , this obviously does not quantify nonclassical correlation of systems described by separable density matrices . instead of negativity , one can define another measure using the partial transposition : @xmath39 where @xmath40 are the eigenvalues of @xmath31 and @xmath41 are the eigenvalues of @xmath42 ; both @xmath40 s and @xmath41 s are aligned in the descending ( or ascending ) order .
this measure utilizes the fact that @xmath43 is eigenvalue - preserving while @xmath44 is , in general , not .
the partial transposition @xmath44 preserves eigenvalues when acting on a density matrix having a biproduct eigenbasis . in this sense , any eigenvalue - preserving - but - not - completely - eigenvalue - preserving map ( ence , this might be reminiscent of pncp ) can be used to define a measure of nonclassical correlation .
this will be investigated in detail in our forthcoming contribution @xcite .
one drawback is that a measure in the form of @xmath45 does not possess an additivity property , similarly to the case of negativity .
the extension of this measure to a multipartite case can be accomplished by taking the minimum over all bipartite splittings .
we compare the four measures we have seen above together with negativity in three simple examples of bipartite splitting cases .
the measure @xmath35 requires a random search of local bases to estimate its value for a given state .
we try @xmath46 randomly generated bases for each data point in figs .
[ fig ] ( b ) and ( c ) .
other measures can be calculated without numerical estimation .
the first example is the pseudo - entangled state of two qubits , @xmath47 with @xmath48 . for this state
, we have @xmath49-s[(1-p)/4]-s[(1 + 3p)/4]$ ] where @xmath50 ( @xmath51 ) . in addition
, we have @xmath52 $ ] where @xmath53 is the binary entropy function .
it is also easy to obtain @xmath54|$ ] and @xmath55 .
these results are plotted against @xmath56 in fig .
[ fig ] ( a ) .
the second example is the two - qubit density matrix @xmath57 with @xmath58 and @xmath59 .
this is inseparable for @xmath60 since @xmath61 has the eigenvalues @xmath62 , @xmath56 ( with the multiplicity of two ) , and @xmath63 .
we need to estimate @xmath64 using a numerical search . as for other measures ,
we obtain @xmath65 , @xmath66 , @xmath67|$ ] , and @xmath68 ( @xmath69 ) , @xmath70 ( @xmath71 ) , @xmath72 ( @xmath73 ) .
these functions are illustrated in fig .
[ fig ] ( b ) as functions of @xmath56 .
the third example is the @xmath74 density matrix @xmath75 of the bipartite system @xmath76 with the dimensions of the hilbert spaces , two for @xmath77 and four for @xmath78 .
@xmath79 \end{split}\ ] ] with @xmath80 .
this was originally introduced by horodecki @xcite as an entangled state with positive partial transpose .
in fact @xmath81 although it is inseparable for @xmath82 .
it is notable that @xmath83 does not change the eigenvalues of @xmath75 and hence @xmath84 .
the values of other measures are numerically estimated or analytically calculated in a straightforward manner while these are too lengthy to include in the text .
plots of the measures against @xmath85 are shown in fig . [
fig ] ( c ) .
+ @xmath86(a ) + @xmath86(b ) + + @xmath86(c )
there are many different ways to define a measure of nonclassical correlation . we have seen four of them .
the measures @xmath35 and @xmath87 look stable and faithful against changes of parameters among tested measures as far as we could see in the three simple examples .
further investigation is required to find desirable properties in addition to additivity properties .
computational cost should be another factor to choose a measure .
a seemingly natural measure @xmath35 can not be used for a system with a large dimension due to the cost of searching over all possible local bases .
thus the measures other than @xmath35 are good candidates in this sense . we have found that a more general framework to detect and quantify nonclassical correlation can be constructed with ence , in analogy with the pncp map theory , in the process of defining the measure @xmath45 .
consider a map @xmath88 such that @xmath88 preserves the eigenvalues of a density matrix while @xmath89 in general does not .
it is obvious that @xmath89 preserves eigenvalues of a density matrix if it has a biproduct eigenbasis .
thus such @xmath88 can be used to detect and quantify nonclassical correlation .
further investigation in this approach will be reported elsewhere @xcite .
as and rr are independently supported by the grant - in - aids for jsps fellows ( no . 1808962 and no .
1907329 ) .
mn is supported by `` open research center '' project for private universities : matching fund subsidy from mext .
he is also supported by the grant - in - aid for scientific research from jsps ( grant no .
19540422 ) .
99 r. f. werner , phys .
a * 40 * ( 1989 ) 4277 .
a. peres , phys .
rev . lett .
* 77 * ( 1996 ) 1413 .
a. peres , quant - ph/9609016 . m. b. plenio and s. virmani , quant .
inf . comp .
* 7 * ( 2007 ) 1 . c. h. bennett , d. p. divincenzo , c. a. fuchs , t. mor , e. rains , p. w. shor , j. a. smolin , and w. k. wootters , phys .
a * 59 * ( 1999 ) 1070 .
h. ollivier and w. h. zurek , phys .
* 88 * ( 2001 ) 017901 .
j. oppenheim , m. horodecki , p. horodecki , and r. horodecki , phys .
* 89 * ( 2002 ) 180402 .
m. horodecki , p. horodecki , r. horodecki , j. oppenheim , a. sen(de ) , u. sen , and b. synak - radtke , phys . rev . a * 71 * ( 2005 ) 062307 .
b. groisman , d. kenigsberg , and t. mor , quant - ph/0703103 .
a. saitoh , r. rahimi , and m. nakahara , quant - ph/0703133 .
k. @xmath90 , p. horodecki , a. sanpera , and m. lewenstein , phys .
a * 58 * ( 1998 ) 883 ; g. vidal and r. f. werner , phys .
a * 65 * ( 2002 ) 032314 . | we introduce and compare several measures of nonclassical correlation defined on the basis of a widely - recognized paradigm claiming that a multipartite system represented by a density matrix having no product eigenbasis possesses nonclassical correlation . #
1#1 : |
according to the boltzmann equation ( [ boltz ] ) , a rarefied gas confined in a bounded domain , in contact with a thermal reservoir modeled by ( [ fdiffuse ] ) at a given constant temperature @xmath11 ( isothermal ) , has an equilibrium state described by the maxwellian @xmath12and it is well known @xcite that such an equilibrium is reached exponentially fast , at least if the initial state is close to the above maxwellian in a suitable norm .
if the temperature @xmath11 at the boundary is not uniform in ( fdiffuse ) , such a statement is not true and the existence of stationary solutions and the rate of convergence require a much more delicate analysis , because rather complex phenomena are involved .
for example , suppose that the domain is just a slab between two parallel plates at fixed temperatures @xmath13 and @xmath14 with @xmath15 .
then , one expects that a stationary solution is reached , where there is a steady flow of heat from the hotter plate to the colder one .
the approach to the stationary solution may involve convective motions , oscillations and possibly more complicated phenomena .
even the description of the stationary solution is not obvious , and the relation between the heat flux and the temperature gradient , ( e.g. the fourier law ( [ fourierlaw1 ] ) ) , is not _ a priori _ known . a first answer to such questions
can be given confining the analysis to the small knudsen number regime ( i.e. , k@xmath16 in ( [ boltzmann ] ) ) , where the particles undergo a large number of collision per unit time and a hydrodynamic regime is established .
in this case , it can be formally shown , using expansion techniques c , cc , that the lowest order in k@xmath17 is a maxwellian local equilibrium and the evolution is ruled by macroscopic equations , such as the navier - stokes equations . in particular , the heat flux vector @xmath18 turns out to be proportional to the gradient of temperature , as predicted by the fourier law @xmath19 with the heat conductivity @xmath20 depending on the interaction potential .
this was first obtained by maxwell and boltzmann @xcite which relates the macroscopic heat flow to the microscopic potential of interaction between the molecules .
the rigorous proof of such a statement was given in @xcite in the case of the slab geometry and provided that @xmath21 is sufficiently small ( uniformly in the knudsen number k@xmath17 ) .
this is a special case of a problem which has received recently a large attention in the statistical mechanics community , the derivation of the fourier law from the microscopic deterministic evolutions ruled by the newton or schrdinger equation or from stochastic models @xcite .
the aim of this paper is to analyze the thermal conduction phenomena in the kinetic regime .
this problem was studied in the slab geometry , for small knudsen numbers in @xcite and for large knudsen number ( k@xmath22 ) in @xcite .
here we are interested in a general domain and in a regime where the knudsen number k@xmath17 is neither small nor large . in this regime ,
the only construction of solutions to ( boltzmann ) and ( [ fdiffuse ] ) we are aware of , was achieved in @xcite in a slab for large @xmath21 , with @xmath23 techniques closer in the spirit to the diperna - lions renormalized solutions @xcite ( see also @xcite and references quoted therein ) .
however , the uniqueness and stability of such @xmath24 solutions are unknown . to develop the quantitative analysis we have in mind ,
the theory of the solutions close to the equilibrium ( @xcite ) is better suited .
for this reason we confine ourselves to temperature profiles at the boundary which do not oscillate too much .
more precisely , we will assume that the temperature @xmath6 on @xmath25 is given by @xmath26 with @xmath27 a small parameter and @xmath28 a prescribed bounded function on @xmath29 such that @xmath30this will allow us to use perturbation arguments in the neighborhood of the equilibrium at the _ _ uniform _ _ temperature @xmath31 .
we shall consider the boltzmann equation @xmath32with @xmath33 the probability density that a particle of the gas at time @xmath34 is in a small cell of the phase space @xmath35 centered at @xmath36 . here
@xmath1 is a bounded domain in @xmath2 , @xmath37 with a smooth boundary @xmath29 .
the function @xmath38 is required to be a positive function on @xmath35 such that @xmath39 is fixed for any @xmath40 .
the right hand side of ( [ boltz ] ) , @xmath41 , is the boltzmann collision operator ( non - symmetric ) @xmath42where @xmath43 with @xmath44 ( hard potential ) , @xmath45 ( angular cutoff ) is the collision cross section and @xmath46 are the incoming velocities in a binary elastic collision with outgoing velocities @xmath47 and impact parameter @xmath48 : @xmath49,\quad v_{\ast } ^{\prime } = v_{\ast } + \omega \lbrack ( v - v_{\ast } ) \cdot \omega ] .
\label{elast}\ ] ] the contact of the gas with thermal reservoirs is described by suitable boundary conditions .
we confine ourselves to the simplest interesting case of the diffuse reflection ( [ fdiffuse ] ) , although more general boundary data could be studied @xcite . on @xmath29 (
supposed to be a @xmath50-smooth surface with external normal @xmath51 well defined in each point @xmath52 ) we assume the condition : @xmath53for @xmath52 and @xmath54 , where @xmath55 is the maxwellian at temperature @xmath11 , @xmath56 , \label{maxw}\]]normalized so that @xmath57 throughout this paper , @xmath1 is a connected and bounded domain in @xmath2 , for @xmath37 and the velocity @xmath58 and @xmath59 such that @xmath60we denote the phase boundary in the phase space @xmath61 as @xmath62 , and split it into the outgoing boundary @xmath63 , the incoming boundary @xmath64 , and the grazing boundary @xmath65 : @xmath66the backward _ exit time _
@xmath67 is defined for @xmath68 @xmath69and @xmath70 .
furthermore , we define the singular grazing boundary @xmath71 , a subset of @xmath65 , as : @xmath72and the discontinuity set in @xmath73 : @xmath74 we will use the short notation @xmath75 for the maxwellian @xmath76^{2}}\exp \left [ -\frac{|v|^{2}}{2[\theta _ { 0}+\delta \vartheta ( x)]}\right ] . \label{mudelta}\]]moreover , to denote the global maxwellian at temperature @xmath77 , @xmath78 , we will simply use the symbol @xmath79 : @xmath80since ( [ norm ] ) is valid for all @xmath27 , we have @xmath81 we denote by @xmath82 the standard linearized boltzmann operator @xmath83=\nu ( v)f - kf \notag \\ & = & \nu ( v)f-\int_{\mathbf{r}^{3}}\mathbf{k}(v , v_*)f(v_*)dv _ * , \label{linbol}\end{aligned}\]]with the collision frequency @xmath84 for @xmath44 .
moreover , we set @xmath85finally , we define @xmath86 thanks to ( [ munorm ] ) , @xmath87 , viewed as function on @xmath88 for any fixed @xmath89 , is a @xmath90-projection with respect to the measure @xmath91 for any boundary function @xmath92 defined on @xmath63 .
we denote @xmath93 either the @xmath94norm or the @xmath95norm in the bulk , while @xmath96 is either the @xmath97norm or the @xmath98norm at the boundary .
also we adopt the vinogradov notation : @xmath99 is equivalent to @xmath100 where @xmath101 is a constant not depending on @xmath102 and @xmath103 .
we subscript this to denote dependence on parameters , thus @xmath104 means @xmath105 .
denote @xmath106 .
our main results are as follows .
[ main1 ] there exists @xmath107 such that for @xmath108 in ( [ mudelta ] ) and for all @xmath109 , there exists a non - negative solution @xmath110 with @xmath111 to the steady problem ( [ boltzmann ] ) and ( [ fdiffuse ] ) such that for all @xmath112 , @xmath113if @xmath114 with @xmath115 is another solution such that , for @xmath116 @xmath117then @xmath118 .
furthermore , if @xmath6 is continuous on @xmath29 then @xmath119 is continuous away from @xmath120 . in particular ,
if @xmath1 is convex then @xmath121 .
on the other hand , if @xmath1 is not convex , we can construct a continuous function @xmath28 on @xmath29 in ( [ temper ] ) with @xmath122 such that the corresponding solution @xmath119 in not continuous .
we stress that the solution @xmath8 is a genuine non equilibrium steady solution .
indeed , it is not a local maxwellian because it would not satisfy equation ( [ boltzmann ] ) , nor a global maxwellian because it would not satisfy the boundary condition ( [ fdiffuse ] ) .
we also remark that in addition to the wellposedness property , the continuity property in this theorem is the first step to understand higher regularity of @xmath119 in a convex domain .
the robust @xmath123 estimates used in the proof enable us to establish the following @xmath124expansion of @xmath119 , which is crucial for deriving the necessary condition of the fourier law ( [ fourierlaw1 ] ) . in the rest of this paper
we assume the normalization @xmath125 .
[ main2 ] let @xmath126 with @xmath127 for all @xmath128 from ( [ norm ] ) , and set @xmath129 then there exist @xmath130 with @xmath131 , for @xmath132 , such that the following @xmath133-expansion is valid @xmath134 with @xmath135 for @xmath112 .
in particular , @xmath136 satisfies @xmath137 we have the following dynamical stability result : [ main3 ] for every fixed @xmath112 , there exist @xmath138 and @xmath139 , depending on @xmath140 , such that if @xmath141\|_{\infty } + |\langle v\rangle ^{\beta } e^{\zeta |v|^{2}}[f(0)-f_{s}]|_{\infty } \leq \varepsilon _ { 0},\label{epsilon0}\ ] ] then there exists a unique non - negative solution @xmath142 to the dynamical problem ( [ boltz ] ) and ( [ diffbc ] ) such that @xmath143\|_{\infty } + |\langle v\rangle ^{\beta } e^{\zeta |v|^{2}}[f(t)-f_{s}]|_{\infty } \\ & & \ \lesssim \
e^{-\lambda t}\big\{\|\langle v\rangle ^{\beta } e^{\zeta convex , @xmath6 is continuous on @xmath29 and moreover @xmath144 is continuous away from @xmath65 and satisfies the compatibility condition @xmath145then @xmath33 is continuous away from @xmath65 .
the asymptotic stability of @xmath119 further justifies the physical importance of such a steady state solution .
we remember that , when @xmath146 , then @xmath147 fails to be a global maxwellian or even a local maxwellian , and its stability analysis marks a drastic departure from relative entropy approach ( e.g. @xcite ) .
moreover , such an asymptotic stability plays a crucial role in our proof of non - negativity of @xmath119 .
.2 cm an important consequence of theorem [ main2 ] is the corollary below which specializes the result to the case of a slab @xmath148 between two parallel plates kept at temperatures @xmath149 and @xmath150 .
[ fourierlaw0]let @xmath151 $ ] and @xmath152 be the solution to @xmath153^{2}}\exp \left [ -\frac{|v|^{2}}{2[1-\delta ] } \right ] \int_{u_{1}<0}u_{1}{\sqrt{\mu(u)}}f_{s}(-\frac{1}{2},u)du,\text { \ \ \ } v_{1}>0 , \notag \\
{ \sqrt{\mu(v)}}f_{s}(\frac{1}{2},v ) & = & \frac{1}{2\pi \lbrack 1+\delta ] ^{2}}\exp \left [ -\frac{|v|^{2}}{2[1+\delta ] } \right ] \int_{u_{1}>0}u_{1}{\sqrt{\mu(u)}}f_{s}(\frac{1}{2},u)du,\text { \ \ \ \ \ \ } v_{1}<0
. \notag\end{aligned}\]]where @xmath154 is the component in the direction @xmath155 of the velocity @xmath156 .
then @xmath136 , the first order in @xmath27 correction to @xmath79 according to the expansion of theorem [ main2 ] , is the unique solution to @xmath157 in order to establish a criterion of validity of the fourier law , let us remember that the temperature associated to the stationary solution @xmath119 is given by @xmath158where @xmath159 and @xmath160 .
the heat flux associated to the distribution @xmath119 is the vector field defined as @xmath161 we state the fourier law in the following formulation : _ there is a positive @xmath50 function @xmath20 , the heat conductivity , such that ( fourierlaw1 ) is valid for _ @xmath119__. _ _ [ fourierlaw ] if the fourier law holds for @xmath119 , then @xmath162 is a linear function over @xmath148 .
available numerical simulations , figure 1 , @xcite indicate the linearity is clearly violated for all finite knudsen number k@xmath17 ( @xmath163 in figure 1 ) .
we therefore predict that the general fourier law ( [ fourierlaw1 ] ) is invalid and inadequate in the kinetic regime with finite knudsen number k@xmath17 .
it is necessary to at least solve the linear boltzmann equation ( [ f1 ] ) to capture the heat transfer in the boltzmann theory .
.3 cm without loss of generality , we may assume @xmath164 and k@xmath165 throughout the rest of the paper .
therefore in the rest of this paper we assume @xmath166 the key difficulty in the study of the steady boltzmann equation lies in the fact that the usual entropic estimate for @xmath167 , coming from @xmath168 , is absent .
the only a - priori estimate is given by the entropy production @xmath169 , which is very hard to use @xcite .
in the context of small perturbation of @xmath79 , only the linearized dissipation rate @xmath170 is controlled , the projection of @xmath92 on the null space of @xmath82 . ] and the key is to estimate the missing hydrodynamic part @xmath171 , in term of @xmath172 .
this is a well - known basic question in the boltzmann theory .
motivated by the studies of collisions in a plasma , a new _ nonlinear energy method _ in high sobolev norms was initiated in the boltzmann study , to estimate @xmath171 ( @xcite ) in terms of @xmath172 . for @xmath173 , remembering the definitions ( [ linbol ] ) of the linearized boltzmann operator and ( [ nonbol ] ) of the quadratic nonlinear collision operator , the boltzmann equation ( boltz ) can be rewritten for the perturbation @xmath92
as@xmath174 _ if _ the operator @xmath82 _ were _ positive definite , then global solutions ( for small @xmath92 ) could be constructed `` easily '' for ( gboltzmann ) .
however , @xmath82 is only semi - positive @xmath175where @xmath176 is the @xmath177-weighted @xmath178 norm .
the kernel ( the hydrodynamic part ) is given by @xmath179note that we use slight different definitions of @xmath180 and @xmath181 from @xcite to capture the crucial total mass constraint .
the so - called ` hydrodynamic part ' of @xmath92 , @xmath171 , is the @xmath90-projection on the kernel of @xmath82 , for every given @xmath155 .
the novelty of such energy method is to show that @xmath82 is indeed _ positive definite _ for small solutions @xmath92 to the nonlinear boltzmann equation ( [ gboltzmann ] ) . the key macroscopic equations for @xmath182 connect @xmath171 and @xmath183 via the boltzmann equation as in page 621 of @xcite:@xmath184where @xmath185 denotes some second order differential operator in @xcite .
such hidden ellipticity in @xmath186 ( @xmath187 ) implies that the hydrodynamic part @xmath171 , missing from the lower bound of @xmath82 , can be controlled via the microscopic part @xmath183 , so that @xmath82 can control the full @xmath188 from ( [ pg ] ) .
such nonlinear energy framework has led to resolutions to several open problems in the boltzmann theory @xcite .
it should be noted that all of these results deal with idealized periodic domains , in which the solutions can remain smooth in @xmath186 for @xmath187 .
of course , a gas is usually confined within a container , and its interaction with the boundary plays a crucial role both from physical and mathematical view points .
mathematically speaking , the phase boundary @xmath189 is always _
characteristic _ but not _ uniformly characteristic _ at the grazing set @xmath190and @xmath191 .
in particular , many of the natural physical boundary conditions create singularities in general domains ( @xcite ) , for which the high sobolev estimates break down in the crucial elliptic estimates ( [ ellipticity ] ) .
discontinuities are expected to be created at the boundary , and then propagate inside a non - convex domain .
therefore completely new tools need to be developed .
a new @xmath178@xmath192framework is developed in @xcite ( see @xcite for a short summary of the method ) to resolve such a difficulty in the boltzmann theory , which leads to resolution of asymptotic stability of maxwellians for specular reflection in an analytic and convex domain , and for diffuse reflection ( with uniform temperature ! ) in general domains ( no convexity is needed ) .
we remark that the non - convex domains occur naturally for non - isothermal boundary ( e.g. two non flat separated boundaries ) .
furthermore , the solutions to the boundary problems are shown to be _ continuous _ if the domain @xmath1 is strictly convex .
different @xmath193 methods have been used in @xcite in particular geometries .
the new @xmath194 framework introduced in @xcite has two parts : @xmath178 _ _ positivity : _ _ assume the wall temperature @xmath11 is constant in ( [ diffbc ] ) .
it suffices to establish the following finite - time estimate@xmath195the natural attempt is to establish @xmath178 estimate for @xmath196 from the macroscopic equation ( [ ellipticity ] ) .
however , this is very challenging due to the fact that only @xmath92 has trace in the sense of green s identity ( lemma [ green ] ) , neither @xmath171 nor @xmath183
@xmath197even make sense on the boundary of a general domain .
instead , the proof of ( [ pi - p ] ) given in @xcite is based on a delicate contradiction argument because it is difficult to estimate @xmath171 via @xmath183 directly in a @xmath178 setting in the elliptic equation ( [ ellipticity ] ) , in the presence of boundary conditions .
the heart of this argument , lies in an exact computation of @xmath171 which leads to the contradiction . as a result , such an indirect method fails to provide a constructive estimate of ( pi - p ) with explicit constants .
@xmath123 _ _ bound : _ _ the method of characteristics can bootstrap the @xmath178 bound into a point - wise bound to close the nonlinear estimate .
let @xmath198 be the semigroup generated by @xmath199 and @xmath200 the semigroup generated by @xmath201 , with the prescribed boundary conditions . by _ two iterations
, _ one can establish : @xmath202from the compactness property of @xmath203 , the main contribution in ( [ duhamel2 ] ) is roughly @xmath204where @xmath205 denotes the generalized characteristics associated with specific boundary condition .
a change of variable from @xmath206 to @xmath207 would transform the @xmath206 and @xmath208-integration in ( [ k - k ] ) into @xmath155 and @xmath156 integral of @xmath92 , which decays from the @xmath178 theory .
the key is to check if @xmath209is valid . without boundary , @xmath210 is simply @xmath211 , and @xmath212 most of the time . for specular or diffuse reflections , each type of characteristic trajectories @xmath213 repeatedly interact with the boundary . to justify ( [ vidav ] )
, various delicate arguments were invented to overcome different difficulties , and analytic boundary and convexity are needed for the specular case . since its inception , this new @xmath194 approach has already led to new results in the study of relativistic boltzmann equation @xcite , in hydrodynamic limits of the boltzmann theory(@xcite
@xcite @xcite ) , in stability in the presence of a large external field ( @xcite @xcite ) . our current study of non - isothermal boundary ( @xmath11 is non - uniform )
is naturally based on such a @xmath214framework .
the main new difficulty in contrast to @xcite , however , is that the presence of a non - constant temperature creates non - homogeneous terms in both the linear boltzmann ( steady and unsteady ) as well as in the boundary condition , so that the exact computation , crucial to the @xmath178 estimate ( [ pi - p ] ) , breaks down , and the @xmath194 scheme @xcite collapses .
the main technical advance in this paper is the development of a _ direct _ ( constructive ) and robust approach to establish ( [ pi - p ] ) in the presence of a non - uniform temperature diffuse boundary condition ( [ fdiffuse ] ) . instead of using these macrosopic equations ( e.g. ( [ ellipticity ] ) ) ,
whose own meaning is doubtful in a bounded domain , we resort to the basic green s identity for the transport equation and choose proper test functions to recover ellipticity estimates for @xmath215 , @xmath216 , @xmath217 and hence @xmath171 directly . in light of the energy identity ,
@xmath218 is controlled at the boundary @xmath219 , but not @xmath220 in ( [ pgamma ] ) .
the essence of the method is to choose a test function which can eliminate the @xmath220 contribution at the boundary , and to control the @xmath215 , @xmath216 , @xmath217 component of @xmath171 respectively in the bulk _ at the same time_. the choice of the test function for @xmath217 is rather direct : we set @xmath221 for some constant @xmath222 to be determined , with @xmath223 and @xmath224 on @xmath225 . on the other hand ,
the test function for @xmath216 is rather delicate .
in fact , two different sets of test functions have to be constructed:@xmath226 and @xmath227 on @xmath228 .
in particular , a unique constant @xmath222 can be chosen to deduce the estimates for @xmath216 , thanks to the special structure of the transport equation and the diffuse boundary condition .
the choice of the test function for @xmath215 requires special attention .
it turns out that in order to eliminate the contribution @xmath220 , we need to choose the test function @xmath229with @xmath230 and neumann boundary condition@xmath231this is only possible if the total mass of @xmath92 is zero , i.e. , @xmath232this illustrates the importance of the mass constraint , which unfortunately is not valid for the steady problem ( [ boltzmann ] ) and ( [ fdiffuse ] ) .
so we are forced to use a penalization procedure ( see below ) to deal with it .
the key lemma which delivers the basic estimates is lemma [ steadyabc ] .
furthermore , in the dynamical case , also the time derivatives @xmath233 , @xmath234 and @xmath235 need to be controlled in negative sobolev spaces .
this is possible due to the special structure of the boltzmann equation as well as the diffuse boundary condition .
once again , the total mass zero condition ( [ a=0 ] ) is essential .
this is the key to prove the crucial lemma [ dabc ] . even though this new unified procedure can be viewed as a ` weak version ' of the macroscopic equations ( [ ellipticity ] ) , the estimates we obtain via this approach are more general .
for instance , the ( [ ellipticity ] ) was only valid for dimension @xmath236 , but the new estimates are valid for any dimension
. there seems to be a very rich structure in the linear boltzmann equation . to bootstrap such a @xmath178 estimate into a @xmath123 estimate
, we define the stochastic cycles for the generalized characteristic lines interacting with the boundary : [ diffusecycles]fixed any point @xmath237 with @xmath238 , let @xmath239 @xmath240 . for
@xmath241such that @xmath242 , define the @xmath243-component of the back - time cycle as @xmath244set @xmath245define @xmath246 , and let the iterated integral for @xmath247 be defined as @xmath248where @xmath249 is a probability measure .
we note that the @xmath250 s ( @xmath251 ) are all independent variables , and @xmath252 depend on @xmath253 for @xmath254 . however , the phase space @xmath255 implicitly depends on @xmath256 .
our method is to use the vidav s two iterations argument(@xcite ) and estimate the @xmath123-norm along these stochastic cycles with corresponding phase spaces @xmath257 .
the key is to estimate measures of various sets in @xmath258 in lemma [ k ] .
we designed an abstract and unified iteration ( [ iteration ] ) , which is suitable for both steady and unsteady cases .
new precise estimate ( [ rho12 ] ) of non - homogeneous terms resulting from the non - constant temperature are obtained in prop .
linfty . based on a delicate change of variables in lemma [ tbsmooth ] ,
such an estimate is crucial in the proof of formation of singularity for a non - convex domain in theorem [ main1 ] . in order to keep the mass zero condition and to start iterating scheme , it is essential to introduce a penalization @xmath259 to solve the problem@xmath260 the presence of @xmath259 ensures the critical zero mass condition ( a=0 )
. it is important to note that our @xmath123 estimate are intertwined with @xmath178 at every step of the approximations , which ensures the preservation of the continuity for a convex domain .
the continuity properties of our final solutions follows from the @xmath123 limit at every step .
moreover , the proof of continuity away from the singular set @xmath120 in ( [ dset ] ) in a general domain is consequence of a delicate result for @xmath261 in @xcite . to illustrate the subtle nature of our construction
, we remark that for the natural positivity - preserving scheme : @xmath262we are unable to prove the convergence , due to breakdown of @xmath263 , whence the mass zero constraint ( [ a=0 ] ) fails to be satisfied .
consequently , we are unable to prove @xmath264 in our construction . such a positivity
is only proven via the dynamical asymptotic stability of @xmath119 , in which the initial positivity plus the choice of a small time interval are crucial to guarantee the convergence of the analog of ( [ positivescheme ] ) in the dynamical setting .
our estimates are robust and allow us to expand our steady state @xmath119 in terms of @xmath27 , the magnitude of the perturbation .
this leads to the first order precise characterization of @xmath119 by @xmath136 , which satisfies ( [ f1 ] ) .
it should be pointed out the our rather complete study of the non - isothermal boundary for the boltzmann theory for @xmath265 forms a mathematically solid foundation for influential work in applied physics and engineering such as in @xcite@xcite , where the existence of such steady solutions to the boltzmann equation is a starting point , but without mathematical justification .
we expect our solutions as well as our new estimates would lead to many new developments along this direction .
.2 cm the plan of the paper is the following . in next section
we present some background material and in particular a version of the ukai trace theorem and the green identity as well as a new @xmath23 estimate at the boundary in lemma 2.3 .
section 3 is devoted to the construction of @xmath266 solutions to the stationary linearized problem .
in particular , we prove lemma [ steadyabc ] which provides the basic estimate of the @xmath266 norms of @xmath171 in terms of the @xmath266 norm of @xmath172 . in section 4 , after introducing an abstract iteration scheme suitable for proving @xmath267 bounds , we prove , in proposition [ linfty ] , the existence of the solution to the linearized problem in @xmath267 and related bounds . in section 5
we combine the results of the previous sections to construct the stationary solution to the boltzmann equation and discuss its regularity properties .
in particular , we give the proof of the @xmath133-expansion and use it to establish a necessary condition for the validity of the fourier law . in section 6
we extend the @xmath266 estimates to the time dependent problem .
section 7 contains the extension of the @xmath267 estimates to the time dependent problem and the proof of the exponential asymptotic stability of the stationary solution . from this
we then obtain its positivity .
in this section we state basic preliminaries .
first we shall clarify the notations of functional spaces and norms : we use @xmath268 for both of the @xmath269 norm and the @xmath270 norm , and @xmath271 for the standard @xmath272 inner product .
moreover we denote @xmath273 and @xmath274 . for the phase boundary integration
, we define @xmath275 where @xmath276 is the surface measure and define @xmath277 and the corresponding space as @xmath278 .
further @xmath279 and @xmath280 .
we also use @xmath281 . denote @xmath282 . recall that @xmath283 for @xmath37 , and @xmath284 . for @xmath285 ,
the notation @xmath286 means @xmath287 . in the next lemma
, we prove that the trace of @xmath92 is well - defined locally for a certain class of @xmath92 : [ ukai ] define @xmath288 then @xmath289 for all @xmath290 .
recall the notation ( [ v ] ) and the definition of backward exit time @xmath291 and @xmath292 in ( [ backexit ] ) .
in the steady case , from @xcite , page 247 , we have the following identity : @xmath293 assume @xmath294 .
then @xmath295 is an absolutely continuous function of @xmath296 for fixed @xmath36 , so that , by the fundamental theorem of calculus @xmath297\times\gamma_+,\\ f(x , v ) & = & f(x+s^{\prime}\bar{v } , v ) + \int_{0}^{s^{\prime } } v \cdot \nabla_x f(x+\tau \bar{v},v)d\tau\nonumber \\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \text{for } \ ( s^{\prime},x , v)\in[0,t_{\mathbf{b}}(x ,- v)]\times\gamma_-,\end{aligned}\ ] ] and therefore , for both cases , @xmath298 assume that @xmath299 is locally parameterized by @xmath300 so that @xmath301 by the compactness of @xmath302 ( and @xmath299 ) , uniformly in @xmath155 , we have @xmath303 for all @xmath304 .
taking inner product of @xmath305 with @xmath51 , we get @xmath306therefore we deduce @xmath307 if @xmath308 or @xmath309 then , by the definition of @xmath310 and ( 40 ) in @xcite , @xmath311 first we integrate ( [ ukai_1 ] ) , for @xmath312 , and then for @xmath313 $ ]
. similarly integrate ( [ ukai_2 ] ) , for @xmath314 , and then for @xmath315 $ ] to have @xmath316 in the dynamic case , from @xcite in page 247 , we have the following identity : @xmath317\times\omega\times\mathbf{r}^3 } f(s , x , v ) ds dx dv = \int_{\gamma_+ } \int_{\max\{0,t - t_{\mathbf{b}}(x , v)\}}^{t } f(s , x-(t - s)\bar{v},v ) ds d\gamma\nonumber\\ & & \hskip 3.5cm+ \int_{\gamma_- } \int_{0}^{\min\{t , t_{\mathbf{b}}(x ,-
v)\ } } f(s , x+s\bar{v},v ) ds d\gamma .
\ \ \ \ \ \ \ \ \ \ \ \ \label{ukai_int}\end{aligned}\ ] ] assume again @xmath318 .
tthen @xmath319 is absolutely continuous of @xmath296 for fixed @xmath320 , so that , by the fundamental theorem of calculus @xmath321\times [ \max\{0,s - t_{\mathbf{b}}(x , v)\},s]\times\gamma_+ , \nonumber\\ f(s , x , v)&= & f(s+s^{\prime},x+ s^{\prime } \bar{v},v ) + \int^{s+s^{\prime}}_{s } v\cdot\nabla_x f ( \tau , x+\tau \bar{v},v ) d\tau \ \ \nonumber\\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \text{for } \ ( s , s^{\prime},x , v)\in[0,t]\times [ 0,\min\{t - s , t_{\mathbf{b}}(x ,- v ) \}]\times\gamma_- , \nonumber\end{aligned}\ ] ] and hence , for both cases , @xmath322for the steady case , assume that @xmath323 , @xmath324 and @xmath325
. then @xmath326 for the dynamic case , assume that @xmath327;l^2(\omega\times\mathbf{r}^3))$ ] , @xmath328 , @xmath329 \times \omega\times\mathbf{r}^3)$ ] and @xmath330\times\partial\omega\times\mathbf{r}^3)$ ] .
then , for almost all @xmath331 $ ] , @xmath332 see the proof in chapter 9 of @xcite .
.3 cm [ tbsmooth ] let @xmath333 and recall the notation ( [ v ] ) .
if @xmath334 then @xmath335 and @xmath336 are smooth functions of @xmath337 so that @xmath338 for @xmath339 , if @xmath340 @xmath341 and if @xmath342 and @xmath343 @xmath344 assume that @xmath299 is locally parameterized by @xmath345 and @xmath302 is locally @xmath346 so that @xmath347by the implicit function theorem , taking @xmath348 and @xmath250 derivative respectively , @xmath349we then have @xmath350 now we prove ( [ change1 ] ) for @xmath351 . without loss of generality
we may assume that @xmath352 .
using the spherical coordinates , @xmath353 and @xmath354 by the implicit function theorem we compute @xmath355 since the jacobian matrix is @xmath356 we have @xmath357 and hence @xmath358 where @xmath359 .
now we apply a change of variables @xmath360 and use ( [ det1 ] ) to further bound ( [ xb ] ) as @xmath361 notice the surface measure of @xmath299 is @xmath362 if @xmath363 .
denote @xmath364 and we use ( [ tlower ] ) as @xmath365 to bound ( [ xb1 ] ) by @xmath366 for @xmath367 , using @xmath368 if @xmath369 we can compute @xmath370 and @xmath371 the rest of proof of ( [ change1 ] ) is same ( even simpler ! ) as the @xmath351 case . for ( [ change2 ] )
, since there is a lower bound for @xmath372 for @xmath373 and @xmath374 , using ( [ det1 ] ) and ( [ det2 ] ) , it is easy to prove ( [ change2 ] ) .
the main purpose of this section is to prove the following : [ linearl2 ] assume @xmath375then there exists a unique solution to @xmath376such that @xmath377 and @xmath378 for the proof of proposition [ linearl2 ] we need several lemmas .
we start with the simple transport equation with a penalization term : [ incoming]for any @xmath379 , there exists a unique solution to @xmath380so that @xmath381 for all @xmath382 .
moreover if @xmath383 and @xmath384 are continuous away from the grazing set @xmath65 , then @xmath92 is continuous away from @xmath120 . in particular ,
if @xmath385 is convex then @xmath386 .
the existence of @xmath92 and @xmath123-bound follow from integration along the characteristic lines of @xmath387 , and @xmath388 ( recall ( [ v ] ) for notations ) .
more precisely , setting @xmath389 , the integrated form of the equation for @xmath390 is : @xmath391 where @xmath67 is defined in ( [ backexit ] ) .
we prove the @xmath392bound by choosing a large @xmath393 such that @xmath394 .
then by the definition of @xmath120 , @xmath395 and hence @xmath67 is smooth by lemma [ tbsmooth ] .
therefore , if @xmath383 and @xmath396 are continuous , @xmath397 is continuous at @xmath398 .
suppose now that @xmath302 is convex . in order to show that @xmath386 , since @xmath399 is true for any @xmath302 , it suffices to show that @xmath400 . since @xmath302 is convex @xmath401 for @xmath402
. therefore @xmath403 .
hence , if @xmath404 then @xmath405 and @xmath406 and @xmath407 .
the @xmath178-estimate and the uniqueness follow from green s formula , since @xmath408 .
.3 cm in next lemma we add to the penalized transport equation a suitably cut - offed linearized boltzmann operator .
moreover we include a _ reduced _ diffuse reflection boundary condition , with the purpose of setting up a contracting map argument .
we have [ lineareb]for any @xmath409 , and for any integer @xmath410 , there exists a unique solution to @xmath411with @xmath412 the linearized boltzmann operator corresponding to the cut - offed cross section @xmath413 .
moreover , uniformly in @xmath414 , we have @xmath415 finally the limit @xmath92 as @xmath416 of the sequence @xmath417 exists and solves uniquely @xmath418 denote @xmath419 . for any @xmath414 , we apply lemma [ incoming ] to the following double - iteration in both @xmath414 and @xmath420 : @xmath421 with @xmath422 . .
we first fix @xmath423 and @xmath414 and take @xmath424 . from green s identity , @xmath425from @xmath426
, we deduce @xmath427moreover , there is @xmath428 such that @xmath429we derive from @xmath430 and @xmath431 , @xmath432since @xmath433 and @xmath434 , by iteration over @xmath420 , for some @xmath435 @xmath436taking the difference of @xmath437 we conclude that @xmath438 is a cauchy sequence .
we take @xmath424 to obtain @xmath439 as a solution to the equation@xmath440 .
we take @xmath441 for @xmath439 . by green s identity
we obtain uniformly in @xmath414 , @xmath442we rewrite for any @xmath443 @xmath444so that from @xmath445 and from the spectral gap of @xmath412 , we have @xmath446but from the equation we have @xmath447^{2}=-2\varepsilon \lbrack f^{j}]^{2}-2f^{j}l_m(\mathbf{i}-\mathbf{p})f^{j}+2f^{j}g.\]]taking absolute value and integrating on @xmath448 , from ( [ epsilonenergy ] ) we have @xmath449hence , by lemma [ ukai ] , for any @xmath450 in ( [ ukaitrace ] ) away from @xmath451 we have @xmath452from ( [ pgamma ] ) we can write @xmath453 for a suitable function @xmath454 and , from @xmath455 , for @xmath456 small @xmath457 where we used the fact @xmath458 therefore we conclude @xmath459adding @xmath460 to @xmath461 , we obtain:@xmath462choosing @xmath463 small and taking the weak limit @xmath464 , we complete the proof of the lemma .
.3 cm next lemma states the crucial @xmath266 bound for @xmath465 .
it will provide uniform in @xmath466 estimates which allow to take the limit @xmath467 in ( [ elinear1 ] ) .
[ steadyabc]let @xmath92 be a solution , in the sense of ( [ weakformulation ] ) below , to @xmath468with @xmath469then we have @xmath470 .3 cm the green s identity ( [ steadygreen ] ) provides the following weak version of ( [ linearf2 ] ) : @xmath471recall that @xmath472\}\sqrt{\mu } $ ] on @xmath448 .
the key of the proof is to choose suitable @xmath473 test functions @xmath474 to estimate @xmath475 and @xmath217 in ( [ cestimate ] ) , ( [ stiijjkiki ] ) , ( [ stijijiijj ] ) , ( [ aestimate ] ) thus concluding the proof of proposition [ linearl2 ] . _
step 1_. estimate of @xmath476 to estimate @xmath217 , we first choose the test function @xmath477where @xmath478and @xmath479 is a constant to be determined . from the standard elliptic estimate , we have @xmath480with the choice ( [ phic ] ) , the right hand side of ( [ weakformulation ] ) is bounded by@xmath481 we have @xmath482so that the left hand side of ( [ weakformulation ] ) takes the form , for @xmath483 @xmath484we decompose @xmath485\big\}\sqrt{\mu } + ( \mathbf{i}-\mathbf{p})f,\text { \ \ \ \ \ on } \omega\times\mathbf{r}^3
. \label{insidesplit}\end{aligned}\]]we shall choose @xmath486 so that , for all @xmath128 @xmath487since @xmath488 the desired value of @xmath479 is @xmath489 . because of the choice of @xmath479 , there is no @xmath215 contribution in the bulk and no @xmath220 contribution at the boundary in ( [ lweak ] ) .
therefore , substituting ( [ bsplit ] ) and ( [ insidesplit ] ) into ( lweak ) , since the @xmath216 terms and the off - diagonal @xmath217 terms also vanish by oddness in @xmath156 in the bulk , the left hand side of ( [ weakformulation ] ) becomes@xmath490 \\ & & -\sum_{i=1}^d \int_{\mathbf{r}^3}(|v|^{2}-\beta_c ) v_{i}^{2 } ( \frac{|v|^{2}}{2}-\frac{3}{2})\mu(v)dv \int_{\omega } \partial _ { ii}{\phi _ { c}(x)c(x ) } dx \\ & & -\sum_{i=1}^d \iint_{\omega \times \mathbf{r}^{3}}(|v|^{2}-\beta_c ) v_{i}\sqrt{\mu } ( { v}\cdot \nabla_x ) \partial _ { i}\phi _ { c}(\mathbf{i}-\mathbf{p})f.\end{aligned}\]]for @xmath491 , @xmath492 .
therefore , we obtain from ( [ rweak ] ) @xmath493where we have used the elliptic estimate and the trace estimate : @xmath494since @xmath495 , from ( [ phic ] ) we obtain @xmath496and hence @xmath497 _ step 2 .
_ estimate of @xmath498 we shall establish the estimate of @xmath216 by estimating @xmath499 for all @xmath500 , and @xmath501 for @xmath502 .
we fix @xmath503 .
to estimate @xmath504 we choose as test function in ( [ weakformulation ] ) @xmath505where @xmath506 is a constant to be determined , and @xmath507from the standard elliptic estimate@xmath508hence the right hand side of ( [ weakformulation ] ) is now bounded by@xmath509now substitute ( [ bsplit ] ) and ( [ insidesplit ] ) into the left hand side of ( [ weakformulation ] ) .
note that @xmath510 is odd in @xmath156 , therefore @xmath87 contributions to ( [ weakformulation ] ) vanishes .
moreover , by ( [ insidesplit ] ) , the @xmath511 contributions to ( [ weakformulation ] ) also vanish by oddness . therefore the left hand side of ( [ weakformulation ] ) takes the form @xmath512\mathbf{1}_{\gamma _ { + } } \label{sti_6 } \\ & & -\int_{\omega } { { \int_{\mathbf{r}^{3}}\sum_{l}(v_{i}^{2}-\beta_b ) v_{l}^{2}\mu \partial _
{ lj}\phi _ { b}^{j}(x)b_{l}dv } } dx \label{sti_8 } \\ & & -\iint_{\omega \times \mathbf{r}^{3}}\sum_{l}(v_{i}^{2}-\beta_b ) v_{l}\sqrt{\mu } \partial _ { l j } \phi _ { b}^{j}(x)(\mathbf{i}-\mathbf{p})f .
\notag\end{aligned}\]]furthermore , since @xmath513 we can choose @xmath514 such that for all @xmath515 @xmath516\mu ( v ) dv=
\int_{\mathbf{r}}[v_{1}^{2}-\beta_b ] e^{-\frac{|v_1|^2}{2 } } dv_1=0 .
\label{alpha}\]]we remark that the choice ( [ alpha ] ) also plays a crucial rule in the dynamical estimate ( [ i_7 ] ) .
since @xmath517 , the desired value is @xmath518 .
note that for such chosen @xmath519 and for @xmath520 , by an explicit computation @xmath521e^{-\frac{|v_1|^2}{2}}dv_{1}= 2\sqrt{2\pi}\neq 0.\end{aligned}\]]therefore , ( [ sti_8 ] ) becomes , by ( [ phibj ] ) , @xmath522}_{=0}\int_{\omega}\partial _ { kj}\phi _ { b}^j ( x)b_{k } \\ & & = 2\sqrt{2\pi}\int_{\omega } ( \partial _ { i}\partial _ { j}\delta ^{-1}b_{j})b_{i}.\end{aligned}\]]hence we have the following estimate for all @xmath503 , by ( [ rweakbj ] ) : @xmath523 in order to estimate @xmath524 for @xmath502 , we choose as test function in ( [ weakformulation ] ) @xmath525where @xmath526 is given by ( [ jb ] ) . clearly , the right hand side of ( [ weakformulation ] ) is again bounded by ( [ rweakbj ] ) .
we substitute again ( [ bsplit ] ) and ( [ insidesplit ] ) into the left hand side of ( [ weakformulation ] ) . the @xmath527 contribution and
@xmath528 contributions vanish again due to oddness
. then the left hand side of ( [ weakformulation ] ) becomes @xmath529\mathbf{1}_{\gamma _
{ + } } \label{stij_6 } \\ & & - { \iint_{\omega \times \mathbf{r}^{3}}|v|^{2}v_{i}^{2}v_{j}^{2}\mu \lbrack \partial _ { ij}\phi _ { b}^{i}b_{j}+\partial _ { jj}\phi _ { b}^{i}(x)b_{i } ] } \label{stij_8 } \\ & & -\iint_{\omega \times \mathbf{r}^{3}}|v|^{2}v_{i}v_{j}v_{k}\sqrt{\mu } \partial _ { kj}\phi _ { b}^{i}(x)[\mathbf{i}-\mathbf{p}]f . \label{stij_9}\end{aligned}\]]note that ( [ stij_8 ] ) is evaluated as @xmath530furthermore , by @xmath531 , @xmath532 , so that @xmath533 combining ( [ stiijjkiki ] ) , we have the following estimate for @xmath534 @xmath535moreover , by ( [ stiijjkiki ] ) , for @xmath536 @xmath537 combining ( [ stijijiijj ] ) and ( [ stjjjj ] ) , we sum over @xmath538 , to obtain , for all @xmath539 @xmath540 _ step 3
. _ estimate of @xmath541 the estimate for @xmath215 is more delicate because it requires the zero mass condition @xmath542 we choose a test function@xmath543where @xmath544it follows from the elliptic estimate with @xmath545 that we have @xmath546since @xmath547 , we can choose @xmath548 so that , for all @xmath515 @xmath549since @xmath550 , the desired value is @xmath551 .
plugging @xmath552 into ( [ weakformulation ] ) and its right hand side is again bounded by@xmath553by ( [ bsplit ] ) and ( [ insidesplit ] ) , since the @xmath217 contribution vanishes in ( [ weakformulation ] ) due to our choice of @xmath554 and the @xmath216 contribution vanishes in ( [ weakformulation ] ) due to the oddness , the right hand side of ( [ weakformulation ] ) takes the form of @xmath555 \ \ \ \ \ \ \label{aboundary } \\ & & -\sum_{i , k=1}^d \iint_{\omega\times\mathbf{r}^3 } ( |v|^{2}-\beta_a ) v_{i}v_{k}\partial _ { ik}\phi _ { a}(x)a(x)\mu(v ) \label{abulk}\\ & & -\sum_{i , k=1}^d\iint_{\omega\times\mathbf{r}^3 } ( |v|^{2}-\beta_a ) v_{i}v_{k}\partial _ { ik}\phi _ { a}(x)(\mathbf{i}-\mathbf{p})f.\label{a2bulk}\end{aligned}\]]we make an orthogonal decomposition at the boundary , @xmath556the contribution of @xmath557 in ( [ aboundary ] ) is @xmath558the crucial choice of ( [ betaalpha ] ) makes the first term vanish due to the neumann boundary condition , while the second term also vanishes due to the oddness of @xmath559 for all @xmath128 .
therefore , ( [ aboundary ] ) and ( [ a2bulk ] ) are bounded by@xmath560the second term ( [ abulk ] ) , for @xmath561 vanishes due to the oddness .
hence we only have the @xmath562 contribution : @xmath563using @xmath564 we obtain @xmath565 we close this section by proving proposition [ linearl2 ] . we keep @xmath423 fixed and take @xmath566 in lemma [ lineareb ] by using lemma [ steadyabc ] which obviously holds even with the additional penalization term . indeed @xmath567 .
hence we have @xmath568 and therefore , for any @xmath569 , @xmath570 we first obtain@xmath571on the other hand from green s identity : @xmath572we deduce from the spectral gap of @xmath412 @xmath573+c_{\eta } [ |r|_{2}^{2}+\|g\|_{2}^{2 } ] . \label{energye}\]]from the argument of ( [ pgamma ] ) and the trace theorem as well as the equation ( [ linearf ] ) , @xmath574plugging this into ( [ energye ] ) with @xmath463 small and adding a small constant @xmath575([pfe ] ) , collecting terms and using the fact @xmath576 we obtain the uniform in @xmath259 estimate @xmath577we thus obtain a weak solution @xmath578 with the same bound ( [ ebound ] ) .
moreover , we have@xmath579+v\cdot \nabla _ { x}[f^{\varepsilon } -f]+l_{m } [ f^{\varepsilon } -f]=\varepsilon f,\text { \ \ \ \ \ \ \ \ } [ f^{\varepsilon } -f ] _ { { -}}=p_{\gamma } [ f^{\varepsilon } -f ] .\]]we conclude from ( [ ebound ] ) , written for the difference of @xmath580 , that @xmath581 |_{2}^{2}&\lesssim & \varepsilon \| f \|_2 ^ 2 \\ & \lesssim & \varepsilon \big\ { \|g\|_{2}^{2}+\|r\|_{2}^{2}\big\}\rightarrow 0 .
\label{ecauchy}\end{aligned}\]]the proposition follows as @xmath566 . finally from ( [ ebound ] ) again
we can take the limit @xmath582 of @xmath583 solution to @xmath584 by a diagonal process , there exists a weak solution @xmath92 such that @xmath585 @xmath197weakly in @xmath586 , for any fixed @xmath587 .
it thus follows from weak semi - continuity of the norm @xmath588 that @xmath589note that the limiting @xmath92 so obtained satisfies @xmath590 the proposition follows as @xmath591 and uniqueness follows from green s identity .
we remark that due to lack of moments control of @xmath92 we can not show @xmath592 strongly in @xmath266 .
we define a weight function scaled with parameter @xmath594 @xmath595 the main purpose of this section is to prove the following : [ linfty ] assume ( [ constraint ] ) .
then the solution @xmath92 to the linear boltzmann equation ( [ linearf ] ) satisfies @xmath596moreover if @xmath383 and @xmath384 are continuous away from the grazing set @xmath65 , then @xmath92 is continuous away from @xmath120 . in particular ,
if @xmath385 is convex then @xmath386 .
we define @xmath597 with a probability measure @xmath598 on @xmath599 which is given by @xmath600 we use below definition [ diffusecycles ] of stochastic cycles and iterated integral and remind the dependence of @xmath601 on @xmath602 . we first show that the set of points in the phase space @xmath258 not reaching @xmath603 after @xmath604 bounces is small when @xmath604 is large .
[ k ] for @xmath605 sufficiently large , there exist constants @xmath606 independent of @xmath607 , such that for @xmath608 , and all @xmath609\times\overline{\omega}\times\mathbf{r}^3,$ ] @xmath610we also have , for @xmath611,@xmath612 for all @xmath613 .
choosing @xmath614 sufficiently small , we further define non - grazing sets for @xmath615 as@xmath616clearly , by the same argument used in ( [ epsilonprime1 ] ) , @xmath617where @xmath101 is independent of @xmath414 . on the other hand ,
if @xmath618 , then from the definition of diffusive back - time cycle ( [ diffusecycle ] ) , we have @xmath619 . by ( [ tlower ] ) , since @xmath620 , and @xmath621 @xmath622therefore , if @xmath623 , then there can be at most @xmath624 + 1 $ ] number of @xmath625 @xmath626 for @xmath615 .
we therefore have @xmath627 + 1}\int_{\{\text{there are exactly } m\text { of } v_{j_{i}}\in \mathcal{v}_{j_{i}}^{{\mathfrak{z}}},\text { and } k-1-m\text { of } v_{j_{i}}\notin \mathcal{v}_{j_{i}}^{{\mathfrak{z}}}\}}\pi _ { j=1}^{k-1}d\sigma _ { j } \\ & \leq & \sum_{m=1}^{\left [ \frac{c_{\xi } t_{0}}{{\mathfrak{z}}^{3}}\right ] + 1}\binom{k-1}{m}|\sup_{j}\int_{\mathcal{v}_{j}^{{\mathfrak{z}}}}d\sigma _ { j}|^{m}\left\ { \sup_{j}\int_{\mathcal{v}_{j}\setminus \mathcal{v}_{j}^{{\mathfrak{z}}}}d\sigma _ { j}\right\ } ^{k - m-1}.\end{aligned}\]]since @xmath628 is a probability measure , @xmath629 , and @xmath630 } \leq \{c{\mathfrak{z}}\}^{k-2-\left [ \frac{c_{\xi } t_{0}}{{\mathfrak{z}}^{3}}\right ] } .\]]but , from @xmath631 + 1}$ ] , we deduce that @xmath632 } ( k-1)^{\left [ \frac{c_{\xi } t_{0}}{{\mathfrak{z}}^{3}}\right ] + 1}c{\mathfrak{z}}^{k-2-\left [ \frac{c_{\xi } t_{0}}{{\mathfrak{z}}^{3}}\right ] } .\label{kdelta}\]]now let @xmath633 + 1\}$ ] , so that if @xmath634 , ( [ kdelta ] ) can be further majorized by @xmath635 + 1}\\ & & \leq \ \left\ { \frac{2nc_{\omega}t_0}{{\mathfrak{z}}^3 } ( c{\mathfrak{z}})^n\right\}^{\left[\frac{c_{\omega}t_0}{{\mathfrak{z}}^3}\right]+1}\\ & & \leq \ \left\ { c_{n,\omega } t_0 { \mathfrak{z}}^{n-3}\right\}^{\left[\frac{c_{\omega}t_0}{{\mathfrak{z}}^3}\right]+1}.\end{aligned}\ ] ] we choose @xmath636 , so that @xmath637 is small for @xmath638 large and for @xmath639 .
moreover , @xmath640 + 1 \sim c_{n,\omega } t_0^{1+\frac{3}{n-3}},\end{aligned}\ ] ] and @xmath641 , if @xmath638 is large so that we can close our estimate .
finally we choose @xmath642 . for @xmath638
sufficiently large , @xmath643 + 1 \sim ct_0^{5/4}$ ] and @xmath644 + 1\}+2 \sim ct_0^{5/4}$ ] , and ( [ largek ] ) follows .
next , we give the proof of the first estimate in ( [ ktildew ] ) .
the left hand side of ( [ ktildew ] ) is bounded by @xmath645 d\sigma_j \right\ } \leq \left\{1 + \frac{1}{\sqrt{2\pi } } \int_{v_1>0 }
\frac { v_1 e^{-(\frac{1}{4}+\zeta)|v|^2 } } { ( 1+\varrho^2 |v|^2)^{\frac{\beta-1}{2 } } } dv \right\}^{k-1 } \\ & & \leq \
\left\ { 1 + \frac{1 } { \sqrt{2\pi } } \int_{\mathfrak{u}_1>0 } \frac { c_{\zeta } \mathfrak{u}_1}{(1+|\mathfrak{u}|^2)^{\frac{\beta-1}{2 } } } \varrho^{-4}d\mathfrak{u } \right\}^{k-1 } \\ & & \leq \ \left\{1 + \frac{c_{\beta,\zeta}}{\varrho^{4}}\right\}^{k-1},\end{aligned}\ ] ] where we used the change of variables : @xmath646 and the fact @xmath611 . for the second estimate in ( [ ktildew ] ) similarly we have @xmath647 d\sigma_l \leq \left\{1 + \frac{1}{\sqrt{2\pi } } \int_{v_1>0 } \frac { v_1 e^{-(\frac{1}{4}+\zeta)|v|^2 } } { ( 1+\varrho^2 |v|^2)^{\frac{\beta-1}{2 } } } dv \right\ } \\ & & \leq \ \left\ { 1 + \frac{1 } { \sqrt{2\pi } } \int_{\mathfrak{u}_1>0 } \frac { c_{\zeta } \mathfrak{u}_1}{(1+|\mathfrak{u}|^2)^{\frac{\beta-1}{2 } } } \varrho^{-4}d\mathfrak{u } \right\ } \\ & & \leq \ \left\{1 + \frac{c_{\beta,\zeta}}{\varrho^{4}}\right\ } , \end{aligned}\ ] ] where we used the fact that @xmath648 is a probability measure on @xmath649 .
+ denote @xmath650 and @xmath651 .
we present an _ abstract iteration _ scheme which gives a unified way to study both steady and dynamic problem with diffuse boundary condition .
recall @xmath652 and @xmath653 from ( [ v ] ) and define : @xmath654(s , x-(t - s){\bar{v}},v)|ds \notag \\ & & + \mathbf{1}_{t_{1}>0}\int_{t_{1}}^{t}e^{-\nu ( v)(t - s)}|[k_{w_{\varrho}}h^{\ell}+w_{\varrho}g](s , x-(t - s){\bar{v}},v)|ds \ \ \ \ \ \ \ \ \ \ \label{iteration } \\ & & + \mathbf{1}_{t_{1}>0}e^{-\nu ( v)(t - t_{1})}|w_{\varrho}r(t_1,x_{1},v)|+\frac{e^{-\nu ( v)(t - t_{1})}}{\tilde{w_{\varrho}}(v)}\int_{\prod_{j=1}^{k-1}\mathcal{v}_{j}}|h| , \notag\end{aligned}\]]where @xmath655 is bounded by @xmath656(s , x_{l}-(t_{l}-s){\bar{v}}_{l},v_{l})|d\sigma _
{ l}(s)ds \ \ \ \ \ \ \ \ \ \ \label{h2 } \\ & & + \sum_{l=1}^{k-1}\int_{t_{l+1}}^{t_{l}}\mathbf{1}_{\{0<t_{l}\}}|[k_{w_{\varrho}}h^{\ell - l}+w_{\varrho } \
g](s , x_{l}-(t_{l}-s){\bar{v}}_{l},v_{l})|d\sigma _ { l}(s)ds \label{h3 } \\ & & + \sum_{l=1}^{k-1 } \mathbf{1}_{\{0<t_{l}\ } } d\sigma _ { l}^r \label{h4 } \\ & & + \mathbf{1}_{\{0<t_{k}\}}|h^{\ell+1-k}(t_{k},x_{k},v_{k-1})|d\sigma _ { k-1}(t_{k } ) , \label{h5}\end{aligned}\]]and we define @xmath657 for the steady case you can regard the temporal variables @xmath658 as parameters and read @xmath659 .
we used the notation in ( [ v ] ) , @xmath660 again and @xmath661 and @xmath662 where @xmath663 is the spatial dimension so that @xmath283 , for @xmath37 . .2 cm [ iterationlinfty]there exist @xmath664 and @xmath665 such that for all @xmath666 , and for @xmath667 , @xmath668 \notag\\ & & \ \
c \max_{1\leq l\leq 2k}\int_{0}^{t } \left\| \frac{h^{\ell - l}(s)}{w_{\varrho}}\right\|_{2}ds.\label{rho}\end{aligned}\ ] ] furthermore for @xmath669 @xmath670 the estimate ( [ rho12 ] ) is used only in the proof of the singularity formation in theorem [ main1 ] .
we first prove ( [ rho ] ) and then sketch the proof of ( [ rho12 ] ) .
we start with @xmath384-contribution in ( [ iteration ] ) and ( [ h4 ] ) .
clearly the contribution in ( [ iteration ] ) is bounded by @xmath671}|w_{\varrho}(v ) r(t_1,x_1,v)|.\ ] ] since the exponent of @xmath672 is bounded by @xmath673 , from ( [ measure ] ) and lemma [ k ] @xmath674 for sufficiently large @xmath675 and @xmath676 .
we now turn to the @xmath383- contribution in ( [ iteration ] ) , ( [ h2 ] ) and ( [ h3 ] ) .
rewrite @xmath677since the exponent of @xmath678 is bounded by @xmath679 , the @xmath383-contributions in ( [ iteration ] ) , ( [ h2 ] ) and ( [ h3 ] ) are bounded by @xmath680 from @xmath681 the second line of ( [ g ] ) is bounded by @xmath682 with our choice @xmath676 and from lemma [ k ] , the third and forth line of ( [ g ] ) are bounded by @xmath683 where we have chosen sufficiently large @xmath664 such that @xmath684 , @xmath685 therefore , the total @xmath383-contribution is bounded by @xmath686 notice that the exponent in @xmath678 is bounded by @xmath687 and from ( [ varrho ] ) and ( [ tweight ] ) and from lemma k we get @xmath688where we have chosen @xmath689 for sufficiently large @xmath675 but fixed and used ( [ varrho ] ) .
now we obtain an upper bound for ( [ iteration ] ) , ( [ h1 ] ) , ( [ h2 ] ) , ( [ h3 ] ) , ( [ h4 ] ) , ( [ h5 ] ) as @xmath690where @xmath691 denotes @xmath692recall from ( [ diffusecycle ] ) that the back - time cycle from @xmath693 denotes @xmath694 we now iterate ( [ hmain ] ) for @xmath695 times to get the representation for @xmath696 and then plug in @xmath697 to obtain @xmath698 the total contributions of @xmath699 s in ( [ hmain ] ) are obtained via plugging ( [ 411 + ] ) with different @xmath700 s into ( [ hmain ] ) .
since @xmath701 , the summation of all contributions of @xmath699 s leads to the bound @xmath702 ^{\varrho}+\left [ \frac{1}{2}\right ] ^ { \varrho } \right\ } \max_{0\leq l\leq k}a_{\ell - l}(t ) \notag\\ & \lesssim_{\beta,\zeta } & \varrho^{\beta}e^{-\frac{\nu _ { 0}}{2}t}\left [ \varrho \sup_{0\leq s\leq t}\left\ { e^{\frac{\nu _ { 0}}{2}s}|wr(s)|_{\infty } \right\ } + \varrho \sup_{0\leq s\leq t}\left\ { e^{\frac{\nu _ { 0}}{2}s}\left\vert \frac{wg(s)}{\langle v\rangle } \right\vert _ { \infty } \right\ } \right ] \notag \\ & & + e^{-\frac{\nu _ { 0}}{2}t}\left\ { \max_{1\leq l\leq 2k}\|h^{\ell+1-l}(0)\|_{\infty } + \left\ { \frac{1}{2}\right\ } ^{\varrho } \max_{0\leq l \leq k}\sup_{0\leq s\leq t}e^{\frac{\nu _ { 0}}{2}s}\|h^{\ell - l+1-k}(s)\|_{\infty } \right\}.\notag\end{aligned}\ ] ] to estimate the @xmath703 contribution , we first separate @xmath704 and @xmath705 . in the first case
, we use the fact @xmath706 and by ( [ 411 + ] ) to obtain the small contribution @xmath707 now we treat the case of @xmath708 . for any large @xmath709 , we can choose a number @xmath710 to define @xmath711 such that @xmath712we split @xmath713 , and the first difference leads to a small contribution in ( [ 411 + ] ) with @xmath714@xmath715 for the remaining main contribution of @xmath716 , note that @xmath717we may make a change of variable @xmath718 ( @xmath719 does not depend on @xmath720 ) so that @xmath721 , for @xmath722 to estimate ( using the notation in ( [ v ] ) ) @xmath723hence the integrand of main contribution is bounded by @xmath724 . rearranging ( [ 411 + ] ) and combining ( [ ant ] ) , ( [ epsilon1 ] ) and ( [ epsilon2 ] ) we have a bound for ( [ 411 + ] ) as @xmath725 \max_{1\leq l\leq 2k}\sup_{0\leq s_1\leq s}\{e^{\frac{\nu _
{ 0}}{2}s_1}\|h^{\ell - l}(s_1)\|_{\infty } \ } \notag\\ & & + e^{-\frac{\nu _ { 0}}{2}s}\bigg\ { \varrho \sup_{0\leq s_1\leq s}\left [ e^{\frac{\nu _ { 0}}{2}s_1}\left|\frac{w_{\varrho}r(s_1)}{\langle v \rangle}\right|_{\infty } \right ] + \varrho \sup_{0\leq s_1\leq s}\left [ e^ { \frac{\nu _ { 0}}{2}s_1}\left\vert \frac{w_{\varrho } g(s_1)}{\langle v\rangle } \right\vert _ { \infty } \right]\notag \\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ + \max_{0\leq l\leq 2k}\|h^{\ell+1-l}(0)\|_{\infty } \bigg\ } \notag\\ & & + c_{\varepsilon , m , n}\max_{1\leq l\leq 2k}\int_{0}^{s}\left\|\frac{h^{\ell - l}(s_1)}{w_{\varrho}(v)}\right\|_{2}ds_1 \notag\\ & \equiv & e^{-\frac{\nu _ {
0}}{2}s}b.\notag\end{aligned}\]]by plugging back ( [ an ] ) and ( [ b ] ) into ( [ hmain ] ) , we obtain@xmath726 \max_{1\leq l\leq 2k}\sup_{0\leq s \leq t}\{e^{\frac{\nu _ { 0}}{2}s } \|h^{\ell - l}(s ) \|_{\infty } \ }
\notag\\ & & + e^{-\frac{\nu _ { 0}}{2}t}\bigg\ { { \varrho } \sup_{0\leq s \leq t}\left [ e^{\frac{\nu _ { 0}}{2}s }
\left|\frac{w_{\varrho}r(s ) } { \langle v \rangle}\right|_{\infty } \right ] + \varrho \sup_{0\leq s \leq t}\left [ e^ { \frac{\nu _ { 0}}{2}s } \left\vert \frac{w_{\varrho } g(s ) } { \langle v\rangle } \right\vert _ { \infty } \right ] \notag \\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ + \max_{0\leq l\leq 2k}\|h^{\ell+1-l}(0)\|_{\infty } \bigg\ } \notag\\ & & + c_{\varepsilon , m , n}\max_{1\leq l\leq 2k}\int_{0}^{t}\left\|\frac{h^{\ell - l}(s ) } { w_{\varrho}(v)}\right\|_{2}ds_1 \notag\\\end{aligned}\ ] ] we then deduce our lemma by choosing @xmath727 and letting @xmath728 large and fixed , ( so are @xmath604 and @xmath40 ) , and then choosing @xmath466 sufficiently small and @xmath729 sufficiently large .
it is trivial to rescale to the case @xmath730 with a different constant depending on @xmath731 .
now we sketch the proof of ( [ rho12 ] ) .
since the proof is similar to ( [ rho ] ) we just highlight the differences .
the key is to estimate the @xmath384-contribution ( [ h4 ] ) differently , using its weaker @xmath732 norm . for large @xmath733
we choose @xmath734 such that @xmath735from ( [ measure ] ) and lemma [ k ] , we bound ( [ h4 ] ) by splitting @xmath736 : @xmath737we have used the @xmath738 is a probability measure for @xmath739 using the notation ( [ v ] ) , @xmath740 and @xmath741 , and we have an uppder bound : @xmath742since @xmath743 and @xmath744 now apply ( [ change1 ] ) in lemma tbsmooth to have @xmath745 we now turn to ( [ hmain ] ) .
now instead of ( [ ant ] ) , combining ( r1l2 ) and ( [ rsplit ] ) yields @xmath746we now plug ( [ diff2 ] ) back into ( [ diff3 ] ) , the estimate for ( [ 411 + ] ) , for each @xmath420 . now we obtain the upper bound for the last term of ( [ diff3 ] ) using the estimate of @xmath747 from ( [ diff2 ] ) as @xmath748where @xmath749is the exit time of the point @xmath750 and @xmath751 we further plug this bound back into ( [ 411 + ] ) and then back into ( hmain ) to isolate _ only _ the contribution of @xmath752 . following exactly the same steps , it suffices to control @xmath753^{\varrho } + \frac{1}{2^{\varrho}}\right\ } d , \notag\end{aligned}\]]by lemma [ k ] . note @xmath754^{\varrho } + \frac{1}{2^{\varrho } } \right\ } \lesssim 1.$ ] we shall use a sequence of approximations to estimate the above integrals . let @xmath755 we again split for @xmath423 large @xmath756we therefore bound@xmath757since @xmath758 is bounded , and @xmath759 we further split @xmath760 and @xmath761 since @xmath762 up to a constant of @xmath763 the main integrals in ( [ ksplit ] ) are bounded by:@xmath764where@xmath765lastly , for any @xmath766 we further split the time intervals of the main integrals in ( [ nsplit ] ) to obtain:@xmath767we now are ready to use change of variables to estimate the main @xmath768-integrals in ( [ esplit ] ) .
recall that , by the definition ( [ primeprime ] ) of @xmath769 and definition ( [ diffusecycles ] ) , @xmath770 reaches @xmath25 if and only if @xmath296 at these @xmath771 from our splitting of time intervals , there exists @xmath772 such that @xmath773 in the integrals in ( [ esplit ] ) , uniformly in @xmath774 we now repeat the change of variables @xmath775 as ( [ change ] ) and ( [ r1l2 ] ) , but using ( [ change2 ] ) instead of ( [ change1 ] ) .
we finally bound ( [ esplit ] ) by@xmath776collecting and combining ( [ diff2 ] ) , ( [ hrmain ] ) , ( [ ksplit ] ) , ( esplit ) and ( [ rchange ] ) , we conclude the @xmath384 contribution in ( hmain ) is bounded by@xmath777we deduce ( [ rho12 ] ) by recalling @xmath778 defined in ( [ diff2 ] ) . after preparing the above tools we return to the stationary problem to give the proof of proposition [ linfty ] .
.2 cm we use the exactly same approximation ( [ approximate ] ) to establish the proposition and follow the same steps in the proof of proposition [ linearl2 ] .
we denote @xmath779 , and rewrite ( approximate ) as@xmath780 _ step 1 _ : we take @xmath424 in @xmath123 . upon integrating over the characteristic lines @xmath781 , and @xmath782 and using the boundary condition repeatedly ,
we obtain ( by replacing @xmath783 with @xmath784 and @xmath177 with @xmath785 ) that the abstract iteration ( [ iteration ] ) is valid for stationary @xmath786 and @xmath787 .
therefore , for @xmath788 , by lemma [ iterationlinfty ] , we get ( choosing @xmath789 large ) @xmath790 + c_k \max_{1\leq l\leq 2k}\left\| f^{\ell - l } \right\|_{2}.\end{aligned}\ ] ] then , absorbing @xmath791 in the left hand side , we have @xmath792 \\ & & + c_{k}\max_{1\leq l\leq 2k}\left\| f^{\ell - l } \right\|_{2}.\end{aligned}\]]now
this is valid for all @xmath788 . by induction on @xmath420 , we can iterate such bound for @xmath793 to obtain@xmath794 \\ & \leq & \frac{1}{4}\max_{1\leq l\leq 2k}\{\|h^{\ell-1+i - l}\|_{\infty } \}+2c_{k}\left [ |w_{\varrho } \
r|_{\infty } + \left\vert \frac{w_{\varrho } \
g}{\langle v\rangle } \right\vert _ { \infty } + \max_{-2k\leq l\leq 2k}\left\| f^{\ell+l } \right\|_{2}\right ] \\ & \vdots & \\ & \leq & \frac{1}{4}\max_{1\leq l\leq 2k}\{\|h^{\ell - l}\|_{\infty } \}+(2i+1)c_{k}\left [ |w_{\varrho } \
r|_{\infty } + \left\vert \frac{w_{\varrho } \
g}{\langle v\rangle } \right\vert _ { \infty } + \max_{-2k\leq l\leq 2k}\left\| f^{\ell+l } \right\|_{2 } \right].\end{aligned}\]]we now take a maximum over @xmath795 to get @xmath796.\notag\\ \label{hn1l}\end{aligned}\]]this implies that , from lemma [ iterationlinfty ] for all @xmath797 @xmath798 .\label{n1-l}\]]now in order to control @xmath799 , we can use ( iteration ) repeatedly for @xmath800 to obtain , by lemma [ k ] , @xmath801we therefore conclude that , from ( [ n1-l ] ) and ( [ l2k ] ) , @xmath802.\end{aligned}\]]but @xmath803 is bounded by step 1 in the proof of lemma [ lineareb ] .
furthermore for @xmath611 , @xmath804 is bounded by @xmath805 and @xmath806 is bounded by @xmath807 .
hence we have @xmath808.\end{aligned}\]]therefore there exists a limit ( unique ) solution @xmath809 .
furthermore , @xmath390 satisfies ( iteration ) with @xmath810 . by subtracting @xmath811 in ( iteration ) with @xmath812
, we obtain from lemma [ iterationlinfty ] , for @xmath811 , ( choosing @xmath813 large ) @xmath814 where @xmath815 is absorbed in the left hand side .
@xmath816 \\ & \leq & \frac{c_{k}(|w_{\varrho } r|_{\infty } + \sup_{0\leq s\leq t}\left\vert \frac{w_{\varrho } g}{\langle v\rangle } \right\vert _ { \infty } ) } { 4^{\ell/2k}}+c_{k}\left [ \max_{-2k\leq l\leq 2k}\|f^{\ell+l}-f\|_{2}\right ] .\end{aligned}\]]from step 1 of lemma [ lineareb ] , we deduce @xmath817 for @xmath420 large
. .3 cm _ step 2 : _ now we let @xmath464 . we take
@xmath439 to be the solution to ( [ jequation ] ) and integrate along @xmath818 repeatedly .
we establish ( [ iteration ] ) for @xmath819 ( we may replace @xmath820 by @xmath783 and @xmath821 to preserve inequality ) and @xmath822 .
lemma [ iterationlinfty ] implies @xmath823 + c(k)\|f^{j}\|_{2 } , \]]so that @xmath824 + c_{k}\|f^{j}\|_{2}.\]]since @xmath825 is bounded , this implies that @xmath826 is uniformly bounded and we obtain a ( unique ) solution @xmath827 .
taking the difference , we have @xmath828+v\cdot \nabla _ { x}[h^{j}-h]+\nu \lbrack h^{j}-h ] = k_{w}[h^{j}-h ] , \\ & & \ \ \ \ \ \ \ \ h _ { { -}}^{j}-h _ { { - } } = \frac{1}{\tilde{w}_{\varrho}(v)}\int_{n(x)\cdot v^{\prime } > 0}[h^{j}-h](t , x , v^{\prime } ) \tilde{w}_{\varrho}(v^{\prime } ) d\sigma(v^{\prime } ) \\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ -\frac{1}{j}\frac{1}{\tilde{w}_{\varrho}(v)}\int_{n(x)\cdot v^{\prime }
> 0}h ^{j}(t , x , v^{\prime } ) \tilde{w}_{\varrho}(v^{\prime } ) d\sigma(v^{\prime } ) .\end{aligned}\]]we regard @xmath829 .
so lemma [ iterationlinfty ] implies that @xmath830,\]]which goes to zero as @xmath414 to @xmath831 .
we obtained a @xmath123 solution @xmath832 to ( elinear ) . by integrating over the trajectory , ( [ iteration ] )
is valid for @xmath833 replaced by @xmath834 so that from lemma [ iterationlinfty ] @xmath835 + c(k)\|f^{\varepsilon } \|_{2 } , \]]which implies that , from proposition [ linearl2 ] , that @xmath836 is uniformly bounded and we obtain @xmath837 solution to the linear equation .
now we have @xmath838+\nu l_m[h^{\varepsilon } -h ] & = & 0 , \\ h _ { { -}}^{\varepsilon } -h _ { { - } } & = & \frac{1}{\tilde{w}_{\varrho}(v)}\int_{n(x)\cdot v^{\prime } > 0}[h^{\varepsilon } -h](t , x , v^{\prime } ) \tilde{w}_{\varrho}(v^{\prime } ) d\sigma , \end{aligned}\]]so that from lemma [ iterationlinfty]@xmath839which goes to zero . for hard potential kernel
, we recall that in previous section we have constructed an approximating sequence @xmath583 to the equation @xmath840with uniform bound in @xmath178 and a limit @xmath585 weakly in @xmath176 , see step 2 in the proof of proposition [ linearl2 ] .
moreover , we obtain from ( [ iterationlinfty ] ) that @xmath841 is uniformly bounded and so is @xmath842 .
note that @xmath843+l_{m}[f^{m}-f]=[-l+l_{m}]f,\text { \ \ \ \ } [ f^{m}-f ] _
{ -}=p_{\gamma } [ f^{m}-f].\]]it follows from proposition [ linearl2 ] and the boundedness of @xmath842 that @xmath844f\|_{2}\rightarrow 0.\]]now , to show @xmath845 , we apply ( [ iteration ] ) with @xmath846 @xmath847since @xmath848 , the second term goes to zero . in the iteration scheme @xmath849 is continuous away from @xmath120 and
hence @xmath92 is continuous away from @xmath120 .
our construction fails to imply the @xmath850 .
this can only be shown by the dynamical asymptotic stability discussed in section 7 .
_ wellposedness_. we consider the following iterative sequence@xmath851with @xmath422 . note @xmath852 and from @xmath853 @xmath854since @xmath855 , we apply proposition [ linfty ] to get @xmath856
, we deduce @xmath857so that for @xmath27 small , @xmath858upon taking differences , we have @xmath859+v\cdot \nabla _ { x}[f^{\ell+1}-f^{\ell}]+l[f^{\ell+1}-f^{\ell } ] \\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ = \gamma ( f^{\ell}-f^{\ell-1},f^{\ell})+\gamma ( f^{\ell-1},f^{\ell}-f^{\ell-1 } ) , \\ & & f _ { { -}}^{\ell+1}-f _ { { -}}^{\ell } = p_{\gamma } [ f^{\ell+1}-f^{\ell}]+\frac{\mu _ { \delta } -\mu } { \sqrt{\mu } } \int_{n(x)\cdot v>0}[f ^{\ell}-f ^{\ell-1}](n(x)\cdot v ) dv+ \sqrt{\mu } , \end{aligned}\]]and by proposition [ linfty ] again for @xmath860 @xmath861\|_{\infty } + |w_{\varrho}[f ^{\ell+1}-f ^{\ell}]|_{\infty } \lesssim \delta \big\ { \|w_{\varrho}[f^{\ell}-f^{\ell-1}]\|_{\infty } + |w_{\varrho}[f ^{n}-f ^{n-1}]|_{\infty } \big\}.\]]hence @xmath438 is cauchy in @xmath123 and we construct our solution by taking the limit @xmath862 .
uniqueness follows in the standard way .
moreover due to theorem 2 and theorem 3 in @xcite @xmath438 is continuous away from @xmath120 and so is @xmath152 .
moreover , if @xmath1 is convex , then @xmath386 .
.2 cm _ formation of singularities_. now we prove the formation of singularity .
recall that @xmath863 and the wall maxwellian is defined as @xmath864 ^ 2}\exp\left[{-\frac{|v|^{2}}{2[1+\delta \vartheta(x ) ] } } \right]\ , \]]while the global maxwellian is @xmath865 . then @xmath866\mu(v ) \delta \vartheta(x ) \ \big| \notag \\ \lesssim \
\delta^2 |\vartheta|_{\infty}^2[1+|v|^4 ] \exp \big[-\frac{|v|^2}{2(1 + \delta|\vartheta|_{\infty})}\big].\label{mudeltaexp}\end{aligned}\ ] ] for any non - convex domain @xmath867 for @xmath339 , there exists at least one @xmath868 with @xmath869 .
suppose @xmath870 satisfies ( [ nlinearfn ] ) with @xmath871 .
we prove that @xmath390 is discontinuous at @xmath872 by contradiction argument .
assume @xmath390 is continuous at @xmath872 where the magnitude of @xmath156 will be chosen later . by definition @xmath873
we choose @xmath874 to be a continuous function such that @xmath875and mainly zero elsewhere .
then by the continuity assumption we know that the quantities @xmath876 and @xmath877 defined below must coincide : @xmath878 .
@xmath876 is obtained by evaluating @xmath879 through the boundary condition @xmath880on the other hand , the existence of @xmath881 allow us to evaluate @xmath882 along the trajectory and has the expression ( [ iteration ] ) with @xmath883 as in ( [ h1 ] ) ( [ h5 ] ) with @xmath884which are independent of time . since from ( [ mudeltaexp ] ) , @xmath885 and [ linfty ] for the last two estimates .
due to ( [ mudeltaexp ] ) we can use the estimate ( rho12 ) for @xmath877 : @xmath886 now we will show that , for a suitable choice of @xmath887 , @xmath888 , which is a contradiction so that @xmath870 has a discontinuity at @xmath872 .
rewrite @xmath889 for large @xmath156 , by ( [ mudeltaexp ] ) , @xmath890 we have @xmath891\sqrt{\mu } \delta |\vartheta & \geq & w_{\varrho } ( v)[\frac{|v|^{2}}{4}-1]\sqrt{\mu } \delta |\vartheta @xmath892 to conclude that @xmath893w_{\varrho } ( v)\sqrt{\mu ( v)}\big)\delta |\vartheta |_{\infty } \big\}\delta |\vartheta |_{\infty } .\end{aligned}\]]by ( [ 2>1 ] ) , we can find @xmath894 so that @xmath895 sufficiently large so that @xmath896\sqrt{\mu ( v_{0})}-\varrho ^{-4}w_{\varrho } ( v_{0})\sqrt{\mu ( v_{0})}\geq c_{v_{0}}>0.\]]on the other hand in ( [ hb ] ) and ( [ 2>1 ] ) , we choose @xmath27 sufficiently small , and @xmath897 sufficiently large , then @xmath729 sufficiently large to get , then @xmath259 sufficiently small , such that for @xmath898 @xmath899 and @xmath900 we can can make @xmath901 arbitrarily small . in particular , there is a continuous function @xmath874 such that @xmath902 hence @xmath903 and this is a contradiction .
.3 cm we now prove the @xmath904expansion .
since @xmath905 is analytic with respect to @xmath133 we have @xmath906 note that @xmath907 , where @xmath908 is some polynomial .
further we seek a formal expansion @xmath909plugging this into the equation , @xmath910+l[\delta f_{1}+\delta ^{2}f_{2}+ \dots ] \\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
= \gamma ( \delta f_{1}+\delta ^{2}f_{2}+ \dots , \delta f_{1}+\delta ^{2}f_{2}+ \dots ) , \\ & & \lbrack \delta f_{1}+\delta ^{2}f_{2}+ \dots ] _ { { - } } = p_{\gamma } [ \delta f_{1}+\delta ^{2}f_{2}+ \dots ] \\ & & \ \ \ \ \ \ \ \ \ + \frac{[\delta \mu _ { 1}+\delta ^{2}\mu _ { 2}+ \dots]}{\sqrt{\mu } } \int_{n(x)\cdot v>0 } [ \sqrt{\mu } + \delta f_{1}+\delta ^{2}f_{2}+ \dots ] \sqrt{\mu } ( n(x)\cdot v)dv.\end{aligned}\]]we compare the coefficients of power of @xmath133 to get an equation for @xmath911 for @xmath912 , ( assuming @xmath913 ) @xmath914 note that , from the @xmath904expansion , @xmath915 since @xmath916 = 0,\end{aligned}\ ] ] by applying proposition [ linfty ] repeatedly , we can construct @xmath130 inductively so that for @xmath917 , @xmath918 in particular @xmath136 satisfies ( [ f1 ] ) .
now we define the remainder @xmath919 satisfying @xmath920 and we obtain the equation for the remainder @xmath919 @xmath921where @xmath922 since by ( [ estfi])@xmath923 we can apply proposition [ linfty ] to deduce that@xmath924we therefore conclude that @xmath925 and the expansion is valid . .3 cm we now consider a slab @xmath926 .
we consider the stationary solution @xmath927 satisfying @xmath928for @xmath929 or @xmath930 . from theorem
[ main1 ] , @xmath931 . we claim that @xmath932\times \mathbf{r}^{3})$ ] for any small @xmath379 . in fact , we multiply ( [ fs ] ) by a 1-dimensional spatial smooth cutoff function @xmath933 , @xmath934 so that@xmath935+l[\chi f_{s}]=\gamma ( \chi f_{s},f_{s})-v_{1}f_{s}\chi ^{\prime } .\]]we take one spatial derivative ( [ fs ] ) to get @xmath936&+&l\partial _ { x}[\chi f_{s } ] = \gamma ( \partial _ { x}[\chi f_{s}],f_{s})+\gamma ( \chi f_{s},\partial _ { x}f_{s})-v_{1}\chi ^{\prime } \partial _ { x}f_{s}-v_{1}\chi ^{\prime \prime } f_{s } \\ & = & \gamma ( \partial _ { x}[\chi f_{s}],f_{s})+\gamma ( \chi f_{s},\partial _ { x}f_{s})+\chi ^{\prime } ( lf_{s}-\gamma ( f_{s},f_{s}))-v_{1}\chi ^{\prime \prime } f_{s } \\ & = & \gamma ( \partial _ { x}[\chi f_{s}],f_{s})+\gamma ( f_{s},\partial _ { x}[\chi f_{s } ] ) \\ & & -\gamma ( f_{s},\chi ^{\prime } f_{s})+\chi ^{\prime } ( lf_{s}-\gamma ( f_{s},f_{s}))-v_{1}\chi ^{\prime \prime } f_{s},\end{aligned}\]]where we have used ( [ fs ] ) to replace @xmath937 . letting @xmath938 , then we have @xmath939note that @xmath940 .
multiply by @xmath941 and use green s identity to have ( no boundary contribution ) @xmath942then we repeat the proof of lemma [ steadyabc ] but replacing @xmath943 in ( [ phia ] ) with the solution of @xmath230 with @xmath944 on @xmath225 .
since the boundary condition of @xmath941 is @xmath945 , lemma [ steadyabc ] is valid so that @xmath946 and @xmath947 is locally in @xmath948 .
recall that @xmath949 now we are going to show @xmath950.\ ] ] for any smooth test function @xmath951)$ ] , due to the weak formulation for @xmath8 , we have for the first component of @xmath952 that @xmath953 where we used the boundary condition as well as the orthogonality of @xmath954 to the collision invariants . hence @xmath955=0 $ ] in the distribution sense and hence @xmath956 .
therefore @xmath957 is continuous up to the boundary and @xmath958 and @xmath959 .
moreover for all @xmath960)$ ] @xmath961 to conclude @xmath962.\ ] ] on the other hand @xmath963 for small @xmath964 so that @xmath965 is positive .
thus @xmath966 .
the components @xmath967 for @xmath968 vanish by symmetry : @xmath969 where @xmath970 and @xmath971 for @xmath972 , and we have defined @xmath973
. then @xmath974 solves @xmath975 where @xmath340 and @xmath54 . by the uniqueness ,
we conclude @xmath976 and hence @xmath977 almost everywhere .
therefore it follows that ( [ i ] ) vanishes and @xmath978 almost everywhere for @xmath968 .
recall the definition of temperature @xmath979 in ( [ thetas ] ) . since @xmath980\times \mathbf{r}^{3})$ ]
, we have @xmath981\times \mathbf{r}^{3})$ ] so that in the sense of distribution , @xmath982with @xmath983 . since @xmath984 , by green s identity , @xmath985 by the fourier law ( [ fourierlaw1 ] ) we have then , in the sense of distribution @xmath986 therefore , in the sense of distributions @xmath987but if @xmath988 , then @xmath989 , and this implies that @xmath990 is in @xmath991 so that @xmath990 is continuous up to the boundary .
but if @xmath983 we thus deduce that @xmath992 itself is continuous up to the boundary @xmath993 .
we then can rewrite @xmath994by the main theorem , @xmath995 , so that @xmath996and we know that @xmath997 is constant in our construction .
then we have for all @xmath998 @xmath999we therefore have @xmath1000 + o(\delta^2),\\ b&=&k(\theta_0 ) + \delta k^{\prime}(\theta_0 ) \frac{\theta_1(+\frac{1}{2})+\theta_1(-\frac{1}{2 } ) } { 2 } + o(\delta^2).\end{aligned}\]]hence the first order expansion @xmath1001 must to be a linear function : @xmath1002 .\]]here @xmath1003 and from ( [ slab ] ) @xmath136 satisfies ( [ slablinear ] ) .
the main purpose of this section is to prove the following : [ dlinearl2]suppose that for all @xmath1004 @xmath1005 then there exists a unique solution to the problem @xmath1006with @xmath1007 such that for all @xmath1008 , @xmath1009 moreover , there is @xmath1010 such that @xmath1011 [ dabc]assume that @xmath383 and @xmath384 satisfy ( [ dlinearcondition ] ) and @xmath92 satisfies ( [ dlinear ] ) , ( [ dlinearl2bc ] ) and ( [ dlinearcondition1 ] ) .
then there exists a function @xmath200 such that , for all @xmath1012 , @xmath1013moreover , @xmath1014 .
the key of the proof is to use the same choices of test functions ( allowing extra dependence on time ) of ( [ phic ] ) , ( [ phibj ] ) , ( [ phibij ] ) and ( [ phia ] ) and estimate the new contribution @xmath1015 in the time dependent weak formulation @xmath1016we note that , with such choices @xmath1017without loss of generality we give the proof for @xmath1018 . *
remark : * we note that ( [ dlinearcondition ] ) , ( [ dlinear ] ) , ( [ dlinearl2bc ] ) and ( [ dlinearcondition1 ] ) are all invariant under a standard @xmath40-mollification for all @xmath1019 .
the estimates in step 1 to step 3 are obtained via a @xmath40-mollification so that all the functions are smooth in @xmath40 . for the notational simplicity
we do not write explicitly the parameter of the regularization .
_ estimate of @xmath1020 in ( [ timeweak ] ) in the weak formulation(with time integration over @xmath1021 $ ] ) , if we choose the test function @xmath1022 with @xmath1023 dependent only of @xmath155 , then we get ( note that @xmath1024 and @xmath383 against @xmath1025 are zero ) @xmath1026\phi ( x ) = 2\pi\sqrt{2\pi}\int_{t}^{t+\varepsilon } \int_{\omega } ( b\cdot \nabla _ { x})\phi ( x ) + \int_{t}^{t+\varepsilon } \int_{\gamma_- } r \phi\sqrt{\mu},\end{aligned}\]]where @xmath1027 and we have used the splitting ( [ bsplit ] ) and ( [ insidesplit ] ) . taking difference quotient , we obtain , for almost all @xmath40 , @xmath1028 notice that , for @xmath1029 , from ( [ dlinearcondition ] ) the right hand side of the above equation is zero so that , for all @xmath1030,@xmath1031 on the other hand , for all @xmath1032 , we have , by the trace theorem @xmath1033 , @xmath1034therefore we conclude that , for all @xmath1019 , @xmath1035where @xmath1036 is the dual space of @xmath1037 with respect to the dual pair @xmath1038 for @xmath1039 and @xmath1040 . by the standard elliptic theory , we can solve the poisson equation with the neumann boundary condition @xmath1041with the crucial condition @xmath1042 for all @xmath1019 .
notice that @xmath1043 where @xmath943 is defined in ( [ phia ] ) .
moreover we have @xmath1044 therefore we conclude , for almost all @xmath1045 @xmath1046 .3 cm _ step 2 .
_ estimate of @xmath1047 in ( [ timeweak ] ) in ( [ timeweak ] ) we choose a test function @xmath1048 . since @xmath550 , @xmath1049
and we get @xmath1050\phi\\ & = & -\int_{t}^{t+\varepsilon } \int_{\gamma } f\phi v_{i}\sqrt{\mu } + 2\pi\sqrt{2\pi } \int_{t}^{t+\varepsilon } \int_{\omega}\partial _ { i}\phi [ a+ c ] \\ & & + \int_{t}^{t+\varepsilon } \iint_{\omega \times \mathbf{r}^{3}}\sum_{j=1}^{d}v_{j}v_{i}\sqrt{\mu } \partial _ { j}\phi ( \mathbf{i}-\mathbf{p})f + \int_{t}^{t+\varepsilon } \iint_{\omega\times\mathbf{r}^3 } \phi v_i g \sqrt{\mu}.\end{aligned}\]]taking difference quotient , we obtain @xmath1051\\ & & + \frac{1}{2\pi\sqrt{2\pi}}\big\{\iint_{\omega \times \mathbf{r}^{3}}\sum_{j=1}^{d}v_{j}v_{i}\sqrt{\mu } \partial _ { j}\phi(\mathbf{i}-\mathbf{p})f(t)+\iint_{\omega\times\mathbf{r}^3 } \phi v_i g(t ) \sqrt{\mu}\big\}.\end{aligned}\]]for fixed @xmath1004 , we choose @xmath1052 solving @xmath1053notice that @xmath1054 where @xmath526 is defined in ( [ jb ] ) .
the boundary terms vanish because of the dirichlet boundary condition on @xmath1055 .
then we have , for @xmath1056 , @xmath1057where we have used the poincar@xmath1058 inequality . hence , for all @xmath1019 @xmath1059 .3 cm _ step 3 .
_ estimate of @xmath1060 in ( [ timeweak ] ) in the weak formulation , we choose a test function @xmath1061 .
since @xmath1062 and @xmath1063 we get @xmath1064and taking difference quotient , we obtain @xmath1065 for fixed @xmath1004 , we define a test function @xmath1066 where @xmath1067 is defined in ( [ phic ] ) .
the boundary terms vanish because of the dirichlet boundary condition on @xmath1068 .
then we have , for @xmath1019 , @xmath1069notice that @xmath1070 in ( [ phic ] ) .
we follow the same procedure of estimates @xmath1071 and @xmath1072 to have @xmath1073where we have used the poincar@xmath1058 inequality .
finally we have , for all @xmath1019 , @xmath1074 .3 cm _ step 4 .
_ estimate of @xmath1075 contributions in ( [ timeweak ] ) to estimate @xmath217 contribution in ( [ timeweak ] ) , we plug ( [ phic ] ) into ( [ timeweak ] ) to have from ( [ insidesplit ] ) @xmath1076the second line has non - zero contribution only for @xmath1077 which leads to zero by the definition of @xmath486 in ( [ beta ] ) .
we thus have from ( [ -1c ] ) , for @xmath466 small , @xmath1078 combining with ( [ cestimate ] ) in lemma [ steadyabc ] we conclude , for @xmath466 small , @xmath1079 to estimate @xmath216 contribution in ( [ timeweak ] ) , we plug ( [ phibj ] ) into ( [ timeweak ] ) to have from ( [ insidesplit ] ) @xmath1080where we used ( [ alpha ] ) to remove the @xmath215 contribution .
we thus have from ( [ -1b ] ) , @xmath1081next we plug ( [ phibij ] ) into ( [ timeweak ] ) and we have from ( [ -1b ] ) , @xmath1082 \big\ } .
\label{bt2}\end{aligned}\ ] ] combining this with ( [ b_i ] ) in lemma [ steadyabc ] we conclude from ( [ timeweak ] ) that @xmath1083 finally in order to estimate @xmath215 contribution in ( [ timeweak ] ) we plug ( [ phia ] ) for into ( [ timeweak ] ) .
we estimate@xmath1084combining this with lemma [ steadyabc ] we conclude @xmath1085 from ( [ timec ] ) , ( [ timeb ] ) and ( [ timea ] ) , we prove the lemma by choosing @xmath466 sufficiently small .
.3 cm the proof is parallel to the steady case , proof of proposition [ linearl2 ] .
we start with the approximating sequence . with @xmath1086 , @xmath1087 .3
cm _ step 1 .
_ we first show that for fixed @xmath414 , @xmath1088 as @xmath424 .
notice that by the compactness of @xmath203 , @xmath1089+c_{\varepsilon } [ \|f^{\ell+1}\|_{2}^{2}+\|f^{\ell}\|_{2}^{2}].\]]we first take an inner product(using green s identity ) with @xmath1090 in ( [ dapproximate ] ) .
with the same choice of @xmath428 in ( [ cj ] ) and using the boundary condition @xmath1091 \int_{0}^{t}|p_{\gamma } f^{\ell}|_{2,-}^{2}\notag\\ & & + c_{j}\int_{0}^{t}|r|_{2 } ^{2}+c_{\varepsilon } \int_0^t \max_{1\leq i\leq \ell+1}\|f^i(s)\|_2 ^ 2 ds + \int_{0}^{t}\|g(s)\|_{\nu } ^{2}ds+\|f_{0}\|_{2}^{2}. \notag \\ & \leq & \varepsilon\int_{0}^{t}\|f^{\ell}(s)\|_{\nu } ^{2}ds+ \left [ ( 1-\frac{1}{j})^{2}+\frac{1}{j^{2}}\right]\int_{0}^{t}|f^{\ell}(s)|_{2,+}^{2}ds \notag\\ & & + c_{j}\int_{0}^{t}|r|_{2 } ^{2}+c_{\varepsilon } \int_0^t \max_{1\leq i\leq \ell+1}\|f^i(s)\|_2 ^ 2 ds + \int_{0}^{t}\|g(s)\|_{\nu } ^{2}ds+\|f_{0}\|_{2}^{2 } \notag \\ & \leq & \max\left\{\frac{\varepsilon}{1-\varepsilon},\left [ ( 1-\frac{1}{j})^{2}+\frac{1}{j^{2}}\right]\right\ } \left\ { ( 1-\varepsilon)\int_{0}^{t}\|f^{\ell}(s)\|_{\nu } ^{2}ds+ \int_{0}^{t}|f^{\ell}(s)|_{2,+}^{2}ds\right\}\notag\\ & & + c_{j}\int_{0}^{t}|r|_{2 } ^{2}+c_{\varepsilon } \int_0^t \max_{1\leq i\leq \ell+1}\|f^i(s)\|_2 ^ 2 ds + \int_{0}^{t}\|g(s)\|_{\nu } ^{2}ds+\|f_{0}\|_{2}^{2}.\notag\end{aligned}\ ] ] we choose @xmath569 sufficiently small , so that @xmath1092\right\ } \leq 1-\varepsilon$ ] .
now we iterate again to have @xmath1093\left\ { c_{\varepsilon}\int_{0}^{t}|r|_{2 } ^{2}+c_{\varepsilon}\int_{0}^{t}\max_{1\leq i\leq \ell+1}\|f^{i}(s)\|_{2}^{2}ds+\int_{0}^{t}\|g(s)\|_{\nu } ^{2}ds+\|f_{0}\|_{2}^{2}\right\ } \notag \\ & \vdots & \notag \\ & \leq & ( 1-\varepsilon ) ^{\ell+1}\left\{(1- \varepsilon)\int_{0}^{t}\|f^{0}(s)\|_{\nu } ^{2}ds+\int_{0}^{t}|f^{0}|^{2}_{2,+}\right\}\notag\\ & & + \frac{1-(1-\varepsilon)^{\ell+1}}{\varepsilon }
\left\ { c_{\varepsilon}\int_{0}^{t}|r|_{2 } ^{2}+c_{\varepsilon}\int_{0}^{t}\max_{1\leq i\leq \ell+1}\|f^{i}(s)\|_{2}^{2}ds+\int_{0}^{t}\|g(s)\|_{\nu } ^{2}ds+\|f_{0}\|_{2}^{2}\right\ } .
\notag\\ \label{lastest}\end{aligned}\]]we therefore have , from @xmath1094 , @xmath1095by gronwall s lemma we have @xmath1096where @xmath1004 is fixed .
this in turns leads to @xmath1097upon taking the difference , we have @xmath1098+v\cdot \nabla _
{ x}[f^{\ell+1}-f^{\ell}]+\nu \lbrack f^{\ell+1}-f^{\ell}]=k[f^{\ell}-f^{\ell-1}],\text { \ \ \ } \label{dndifference}\]]with @xmath1099(0)\equiv 0 $ ] and @xmath1100 $ ] . applying ( [ lastest ] ) to @xmath1101 yields@xmath1102from ( [ lastest ] ) , this implies , from the gronwall s lemma , that , by taking maximum over @xmath1103 @xmath1104therefore , @xmath438 is a cauchy sequence and @xmath1088 so that @xmath439 is the solution of @xmath1105 .3 cm _ step 2 .
_ we let @xmath441 . upon using green s identity and the boundary condition and ( [ eta ] ) , for any @xmath1106 , @xmath1107note that , from ukai s trace theorem ( lemma [ ukai ] ) and ( [ dlinearj ] ) @xmath1108^{2}+v\cdot \nabla _ { x}[f^{j}]^{2}\|_{1 } \notag \\ & \lesssim & \int_{0}^{t}\|f^{j}(s)\|_{2}^{2}ds+\int_{0}^{t } & \lesssim & \int_{0}^{t}\|f^{j}(s)\|_{2}^{2}ds+\int_{0}^{t } \|(\mathbf{i}-\mathbf{p})f^{j}(s)\|_{\nu } ^{2}+\int_{0}^{t}\|g(s)\|_{\nu } ^{2}ds . \notag\end{aligned}\]]but from ( [ dukai ] ) ,
choosing @xmath463 small , we obtain from ( [ jgreen ] ) and ( [ dukai ] ) that@xmath1109it follows from the gronwall s lemma that @xmath1110 and hence @xmath1111since @xmath1112 @xmath1113 , integrating ( [ tlinearj ] ) from @xmath1114 to @xmath40 and combining with ( [ dukai ] ) we have @xmath1115 add ( [ j1 ] ) and ( [ j2 ] ) together to have @xmath1116by taking a weak limit , we obtain a weak solution @xmath92 to ( [ dlinear ] ) with the same bound ( [ jbound ] ) .
taking difference , we have@xmath1117+v\cdot \nabla _ { x}[f^{j}-f]+l[f^{j}-f]=0,\text { \ \ \ \ } [ f^{j}-f]_- = p_{\gamma } [ f^{j}-f]+\frac{1}{j}p_{\gamma } f^{j},\]]with @xmath1118(0)=0 $ ] .
applying the same estimate ( [ jbound ] ) with @xmath1119 we obtain @xmath1120we thus construct @xmath92 as a @xmath266 solution to ( [ dlinear ] ) .
.2 cm in both step 1 and step 2 , we do not need the zero mass constraint which is instead essential for next step .
_ decay estimate . to conclude our proposition ,
let @xmath1121we multiply ( [ dlinear ] ) by @xmath1122 , so that @xmath1123 satisfies @xmath1124we apply green s identity and ( [ dukai ] ) , ( [ jgreen ] ) together with @xmath1125 to obtain@xmath1126from ( [ dlinear ] ) we know that @xmath1127 applying lemma [ dabc ] to ( [ lineary ] ) , we deduce @xmath1128where @xmath1129 .
multiplying with a small constant@xmath575([py])@xmath1130([ygreen ] ) to obtain , for some @xmath1131 @xmath1132+\varepsilon \left\ { \int_{0}^{t}\|(\mathbf{i}-\mathbf{p})y(s)\|_{\nu } ^{2}+\|\mathbf{p}y(s)\|_{\nu } ^{2}\right\ } + \varepsilon \int_{0}^{t}|(i - p_{\gamma } ) y(s)\|^{2}_{2,+ } \\ & \leq & c\lambda \int_{0}^{t}\|y(s)\|_{2}^{2}+[\|y(0)\|_{2}^{2}-\varepsilon g(0)]+c_{\varepsilon , \lambda } \int_{0}^{t}e^{\lambda s}|r|_{2 } ^{2}+\int_{0}^{t}e^{\lambda s}\|g(s)\|_{2}^{2}ds.\end{aligned}\]]by further choosing @xmath1133 , and @xmath1134 for hard potentials , we conclude our proposition .
to conclude the proof of theorem [ main3 ] we need @xmath267 estimates .
they are provided , in the linear case , by next [ dlinearlinfty ] let @xmath1135 and @xmath1136 .
then the solution @xmath92 to ( [ dlinear ] ) satisfies @xmath1137furthermore , if @xmath1138 , @xmath1139 and @xmath383 are continuous , then @xmath1140 is continuous away from @xmath120 .
in particular , it @xmath302 is convex then @xmath386 .
we use exactly the same approximating sequence and repeat step 1 to step 3 of proposition [ dlinearl2 ] to show that @xmath1141 are bounded , via lemma [ iterationlinfty ] .
we denote @xmath1142 , where @xmath1143 is scaled weight in ( [ weight ] ) . rewrite ( [ dapproximate ] ) as@xmath1144 _ step 1._we take @xmath424 in @xmath123 . upon integrating over the characteristic lines
@xmath781 , and @xmath782 and the boundary condition repeatedly , along the stochastic cycles , we obtain ( by replacing @xmath784 with @xmath783 and @xmath821 ) that ( [ iteration ] ) is valid for @xmath1145 . therefore , for @xmath1146 , by lemma [ iterationlinfty ] that @xmath1147 \\ & \equiv & \frac{1}{8}\max_{0\leq l\leq 2k}\sup_{0\leq s\leq t_0}\{e^{\frac{\nu _ { 0}}{2}s}\|h^{\ell - l}(s)\|_{\infty } \}+c(k)\max_{1\leq l\leq 2k}\int_{0}^{t_0}\|f^{\ell - l}(s)\|_{2}ds \\ & & + d,\end{aligned}\]]where @xmath1148 are chosen to be sufficiently large but fixed
. now this is valid for all @xmath788 . by induction on @xmath420 , we can iterate such bound for @xmath793 to obtain@xmath1149taking maximum over @xmath1150 , and by induction on @xmath420 , @xmath1151+d \\ & \leq & \frac{1}{4^{2}}\max_{1\leq l\leq 2k}\sup_{0\leq s\leq t_0}\{e^{\frac{\nu _ { 0}}{2}s}\|h^{\ell-1-l}(s)\|_{\infty } \}+(1+\frac{1}{4})\big\ { c(k)\big[\max_{-2k\leq l\leq 2k}\int_{0}^{t_0}\|f^{\ell - l}(s)\|_{2}ds\big]+d\big\ } \\ & \vdots & \\ & \leq & \frac{1}{4^{\ell/2k}}\max_{1\leq l\leq 2k}\sup_{0\leq s\leq t_0}\{e^{\frac{\nu _ { 0}}{2}s}\|h^{l}(s)\|_{\infty } \}+\big\ { c(k)\big[\max_{-2k\leq l\leq 2k}\int_{0}^{t_0}\|f^{\ell - l}(s)\|_{2}ds\big]+d\big\},\end{aligned}\]]where we may assume that @xmath420 is a multiple of @xmath604 .
now for @xmath1152 , we can use ( [ iteration ] ) for @xmath1153 repeatedly for @xmath1154 to obtain @xmath1155we therefore conclude that @xmath1156now @xmath803 is bounded by step 1 in the proof of proposition [ dlinearl2 ] , @xmath804 is bounded by @xmath805 for @xmath1157 and @xmath806 is bounded by @xmath1158 .
hence , there is a limit ( unique ) solution @xmath1159 .
furthermore , @xmath390 satisfies ( iteration ) with @xmath810 . by subtracting @xmath811 in ( iteration ) with @xmath1160
, we obtain from lemma [ iterationlinfty ] for @xmath811 and ( dnbound):@xmath1161from step 1 of lemma [ dlinearl2 ] we deduce @xmath1162 for @xmath420 large .
.3 cm _ step 2 .
_ we take @xmath464 .
let @xmath439 to be the solution to ( [ dlinearj ] ) and integrate along @xmath818 repeatedly . again
( [ iteration ] ) is valid for @xmath819 ( replacing @xmath820 by @xmath783 and @xmath821 ) and @xmath822 .
lemma [ iterationlinfty ] implies@xmath1163,\end{aligned}\]]with @xmath1164 , and @xmath731 sufficiently large but fixed .
therefore , by an induction over @xmath414 , @xmath1165.\end{aligned}\]]since @xmath1166 is bounded from step 2 of proposition [ dlinearl2 ] ( with no mass constraint ) , this implies that @xmath826 is uniformly bounded and we obtain a ( unique ) solution @xmath1167 .
taking the difference , we have @xmath1168&+&v\cdot \nabla _
{ x}[h^{j}-h]+\nu [ h^{j}-h ] = k_{w_{\varrho}}[h^{j}-h ] , \\ { [ h^{j}-h]}_- & = & \frac{1}{\tilde{w}_{\varrho}(v)}\int_{n(x)\cdot v^{\prime } > 0}[h ^{j}-h ] ( t , x , v^{\prime } ) \tilde{w}_{\varrho}(v^{\prime } ) d\sigma \\ & & -\frac{1}{j}\frac{1}{\tilde{w}_{\varrho}(v)}\int_{n(x)\cdot v^{\prime } > 0}h ^{j}(t , x , v^{\prime } ) \tilde{w}_{\varrho}(v^{\prime } ) d\sigma .\end{aligned}\]]we regard @xmath1169 as @xmath384 in lemma [ iterationlinfty ] .
so lemma [ iterationlinfty ] implies that@xmath1170which goes to zero as @xmath414 to @xmath831 . to obtain decay estimate ,
we integrate ( [ dlinear ] ) over the trajectory . the estimate
( [ iteration ] ) is valid for @xmath1172 so that from lemma iterationlinfty for @xmath675 sufficiently large we have:@xmath1173 , \notag\end{aligned}\]]where @xmath1174 and @xmath731 sufficiently large .
letting @xmath1175 in ( [ hinfty ] ) , we obtain @xmath1176.\notag\end{aligned}\]]let @xmath1177since @xmath1178 , proposition [ dlinearl2 ] implies that @xmath1179let @xmath423 be an integer .
we look at times @xmath1180 . by induction on @xmath423 , we have from the definition of @xmath1181 and ( [ ht0 ] ) for @xmath1182 @xmath1183t_{0})\|_{\infty } + |h([m+1]t_{0})|_{\infty } \\ & \leq & e^{-\frac{\nu _ { 0}}{2}t_{0}}\|h(mt_{0})\|_{\infty } \\ & & + c(\varrho)e^{-\frac{\nu _ { 0}}{2}t_{0}}\bigg [ \sup_{0\leq s\leq t_{0}}\left\ { |w_{\varrho}r(s+mt_{0})|_{\infty } \right\ } + \sup_{0\leq s\leq t_0}\left\ { \left\vert \frac{w_{\varrho}g(s+mt_{0})}{\langle v\rangle } \right\vert _ { \infty } \right\}\\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ + \int_{0}^{t_{0}}\|f(s+mt_{0})\|_{2}ds\bigg ] \\ & \leq & e^{-\frac{\nu _ { 0}}{2}t_{0}}\|h(mt_{0})\|_{\infty } + c(\varrho)e^{-\frac{\nu _ { 0}}{2}t_{0}-\frac{m\lambda t_{0}}{2}}\times \\ & & \bigg [ \sup_{0\leq s\leq t_{0}}e^{\frac{m\lambda t_{0}}{2}}\left\ { } \\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ + \int_{0}^{t_{0}}e^{\frac{m\lambda t_{0}}{2}}\|f(s+mt_{0})\|_{2}ds\bigg ] \\ & \leq & e^{-\frac{\nu _ { 0}}{2}t_{0}}\|h(mt_{0})\|_{\infty }
+ c(\varrho)(1+t_{0})e^{-\frac{\nu _ { 0}}{2}t_{0}-\frac{m\lambda t_{0}}{2}}r \\ & \leq & e^{-2\frac{\nu _ { 0}}{2}t_{0}}\|h([m-1]t_{0})\|_{\infty } + c(\varrho)(1+t_{0})e^{-\frac{\nu _ { 0}}{2}t_{0}-\frac{m\lambda t_{0}}{2}}r\\ & & + c(\varrho)(1+t_{0})e^{-\frac{\nu _ { 0}}{2}t_{0}}e^{-\frac{\nu _ { 0}}{2}t_{0}}e^{-\frac{[m-1]\lambda t_{0}}{2}}r \\ & = & e^{-2\frac{\nu _ { 0}}{2}t_{0}}\|h([m-1]t_{0})\|_{\infty } + c(\varrho)(1+t_{0})e^{-\frac{\nu _ { 0}}{2}t_{0}-\frac{m\lambda t_{0}}{2}}r[1+e^{-\frac{\nu _ { 0}-\lambda } { 2}t_{0 } } ] \\ &
\leq & e^{-3\frac{\nu _ { 0}}{2}t_{0}}\|h([m-2]t_{0})\|_{\infty } \\ & & + c(\varrho)(1+t_{0})e^{-\frac{\nu _ { 0}}{2}t_{0}-\frac{m\lambda t_{0}}{2}}r[1+e^{-\frac{\nu _ { 0}-\lambda } { 2}t_{0}}+e^{-2\frac{\nu _ { 0}-\lambda } { 2}t_{0 } } ] \\ & \vdots&\\ & \leq & e^{-m\frac{\nu _ { 0}}{2}t_{0}}\|h_{0}\|_{\infty } + c(\varrho)(1+t_{0})e^{-\frac{\nu
_ { 0}}{2}t_{0}-\frac{m\lambda t_{0}}{2}}r\sum_{j}e^{-j\frac{\nu _ { 0}-\lambda } { 2}t_{0 } } \\ & \leq & c_{k,\nu
_ { 0},\lambda,\varrho } e^{-\frac{m\lambda t_{0}}{2}}r.\end{aligned}\]]combining with ( [ hinfty ] ) for @xmath1184 , we deduce that for any @xmath1185 , @xmath1186 .
we deduce our proposition with @xmath1187 for @xmath675 sufficiently large but fixed .
a simple scaling of @xmath731 concludes the proof of the proposition .
we consider the following iteration sequence @xmath1188clearly we have @xmath1189recall @xmath1190 in ( [ weight ] ) .
note that for @xmath1191 @xmath1192 using ( [ mudeltaexp ] ) for @xmath1193 @xmath1194 by proposition [ dlinearlinfty ] , we deduce @xmath1195for @xmath27 small , there exists a @xmath1196 ( uniform in @xmath27 ) such that , if the initial data satisfy ( [ epsilon0 ] ) , then @xmath1197by taking difference @xmath437 , we deduce that@xmath1198+v\cdot \nabla _
{ x}[f^{\ell+1}-f^{\ell}]+l[f^{\ell+1}-f^{\ell } ] \\ & & \ \ \ \ = l_{\sqrt{\mu}f_s}[f^{\ell}-f^{\ell-1}]+\gamma ( f^{\ell}-f^{\ell-1},f^{\ell})+\gamma ( f^{\ell-1},f^{\ell}-f^{\ell-1 } ) , \\ & & [ f^{\ell+1}-f^{\ell}]_- = p_{\gamma } [ f^{\ell+1}-f^{\ell } ] + \frac{\mu _ { \delta } -\mu } { \sqrt{\mu } } \int_{\gamma _ { + } } [ f^{\ell}-f^{\ell-1}](n(x)\cdot v ) dv,\end{aligned}\]]with @xmath1199 initially .
repeating the same argument , we obtain @xmath1200(s)\|_{\infty } + \sup_{0\leq s\leq t}|e^{\frac{\lambda s}{2}}w_{\varrho } [ f ^{\ell+1}-f ^{\ell}](s)|_{\infty } \\ & \lesssim & [ \delta + \sup_{0\leq s\leq t}\|e^{\frac{\lambda s}{2}}w_{\varrho } f^{\ell}(s)\|_{\infty } + \sup_{0\leq s\leq t}\|e^{\frac{\lambda s}{2}}w_{\varrho } f^{\ell-1}(s)\|_{\infty } ] \sup_{0\leq s\leq t}\|e^{\frac{\lambda s}{2}}w_{\varrho } [ f^{\ell}-f^{\ell-1}](s)\|_{\infty } .\end{aligned}\]]this implies that @xmath1090 is a cauchy sequence .
the uniqueness is standard .
.3 cm _ positivity . _ to conclude the positivity of @xmath119 , we need to show @xmath1201if initially @xmath1202 . to this end
, we first assume the cross section @xmath1203 is bounded and hence @xmath1204 is bounded .
we note that previous approximating sequence is not well suited to show the positivity .
therefore , we need to design a different iterative sequence .
we use the following one : @xmath1205where @xmath1206 and @xmath1207 with initial condition @xmath1208 .
clearly , such an iteration preserves the non - negativity . we need to show @xmath1209 is convergent to conclude the non negativity of the ( unique ! ) limit @xmath1210 .
writing @xmath1211 , we have @xmath1212taking inner product ( green s identity ) with @xmath1090 , from ukai s trace theorem , and @xmath1213 ( bounded cross section @xmath1203 ) and the boundary condition we have @xmath1214\int_{0}^{t}|p_{\gamma } f^{\ell}|_{2}^{2 } + \int_0^t \|f^\ell\|_{2}^2\label{1cdelta } \\ & & + [ \delta + \|wf^{\ell}\|_{\infty } + \|wf^{\ell}\|_{\infty } ^2]\int_{0}^{t}\|f^{\ell+1}\|_{2 } ^{2}.\notag\end{aligned}\]]by ( [ dukai ] ) and the equation ( [ n ] ) , assuming @xmath1215 we have @xmath1216\int_{0}^{t}\|f^ { \ell } \|_{2 } ^{2}\notag\\ & \lesssim & \max_{\ell-1\leq l \leq \ell } \int_0^t \| f^l \|_2 ^ 2 ds .\label{pgmax}\end{aligned}\]]splitting @xmath1217 in ( [ 1cdelta ] ) and using ( [ pgmax ] ) we have @xmath1218 \int_0^t |p_{\gamma}f^\ell|_2 ^ 2 ds
+ \max_{\ell-1\leq l \leq \ell } \int_0^t \|f^l\|_2 ^ 2 ds.\end{aligned}\]]then we iterate to obtain @xmath1219 ^ 2\int_0^t |p_{\gamma}f^{\ell-1}|_2 ^ 2 + [ 1+(1-\delta)]\left\ { \|f^0\|_2 ^ 2 + \max_{1\leq l \leq \ell+1 } \int_0^t\|f^l(s)\|_2 ^ 2 \right\}\notag\\ & \vdots & \notag\\ & \leq & ( 1-\delta)^{\ell+1}\int_0^t |p_{\gamma}f^{0}|_2 ^ 2 + \frac{1-(1-\delta)^{\ell+1}}{\delta}\left\ { \|f^0\|_2 ^ 2
+ \max_{1\leq l \leq \ell+1 } \int_0^t \|f^l(s)\|_2 ^ 2 \right\}.\notag\\ \label{2iterate}\end{aligned}\ ] ] taking maximum over @xmath1220 , @xmath1221 by gronwall s lemma , from @xmath1222 @xmath1223uniformly bounded .
we now apply lemma [ iterationlinfty ] with @xmath1224for some @xmath638 large , with @xmath1225 , using the same arguments as in proof of ( [ ht0 ] ) , to get @xmath1226 \\ & & + c(t_0)\max_{1\leq l\leq 2k}\int_{0}^{t_0}\|f^{\ell-1}(s)\|_{2}ds.\end{aligned}\]]where @xmath788 . for @xmath1227small
, we obtain from @xmath1228 @xmath1229hence we obtain@xmath1230and @xmath1231but for @xmath1232 , we use lemma with @xmath1153 to get @xmath1233so that ( [ assumption ] ) is valid as long as @xmath1234 is sufficiently small .
we therefore obtain uniform bound @xmath1235this leads to @xmath1236 .
furthermore , @xmath92 satisfies ( [ n ] ) with @xmath871 .
therefore @xmath1237 satisfies @xmath1238 - k[f^\ell - f]= r^{\prime } , \ \ \ \ [ f^{\ell+1}-f](0)\equiv 0,\\ & & [ f^{\ell+1}-f]_{+ } = p_{\gamma}[f^\ell -f ] + \frac{\mu_{\delta}-\mu}{\sqrt{\mu } } \int_{\gamma_+ } [ f^\ell - f]\sqrt{\mu } ( n(x)\cdot v ) dv,\end{aligned}\ ] ] where @xmath1239 notice that from ( [ maxt0 ] ) @xmath1240 with the same proof of ( [ pgmax ] ) @xmath1241|_2 ^ 2 \lesssim \max_{\ell-1 \leq i \leq \ell } \| f^i -f \|_2 ^ 2 ds.\ ] ] combining the two above estimates we have @xmath1242 \int_0^t |p_{\gamma}[f^\ell - f]|_2 ^
2 ds + \max_{\ell-1 \leq i \leq \ell}\int_0^t \|f^i - f\|_2 ^ 2 ds.\end{aligned}\ ] ] we iterate as for ( [ 2iterate ] ) to obtain @xmath1243 then applying gronwall s lemma and choosing small @xmath1244 we conclude that @xmath1245 is cauchy in @xmath1246 $ ] and @xmath1247 by uniqueness and @xmath1210 for @xmath1248 .
we can repeat this process to show @xmath1249 for all @xmath1056 .
finally , we take @xmath1250 , so that @xmath1251 and hence @xmath1210 .
therefore @xmath1252 .
this completes the proof in the case @xmath1203 bounded .
to remove this limitation , we use a cut - off procedure as before and reduce to the bounded case where previous result holds .
then we pass to the limit in the cut - off using the a priori bounds and uniqueness .
_ we are deeply indebted with prof .
k. aoki for permitting us to use the plots from the paper @xcite .
y. guo thanks c. mouhot for stimulating discussions and he thanks the support of beijing international mathematical center .
r. e. and r. m. and c. k. wish to thank the math department of brown university and icerm where part of this research was performed , for their kind hospitality and support .
_ the research of y. g. is supported in part by nsf # 0905255 and frg grants as well as a chinese nsf grant # 10828103 .
the research of r. e. and r. m. is partially supported by miur and gnfm - indam .
the research of c. k. is supported in part by herchel smith fund .
arkeryd , l. ; nouri , a. : _ asymptotic techniques for kinetic problems of boltzmann type_. proceedings of the 3rd edition of the summer school in methods and models of kinetic theory " , riv . mat .
parma . 7 ( 2007 ) , 174 .
ohwada , t. ; aoki , k. ; sone y. : _ heat transfer and temperature distribution in a rarefied gas between two parallel plates with different temperatures _
: numerical analysis of the boltzmann equation for a hard sphere molecule , in rarefied gas dynamics : theoretical and computational techniques , edited by e. p. muntz , d. p. weaver , and d. h. campbell ( aiaa , washington , 1989 ) | in the study of the heat transfer in the boltzmann theory , the basic problem is to construct solutions to the following steady problem : @xmath0where @xmath1 is a bounded domain in @xmath2 , @xmath3 , @xmath4 is the knudsen number and @xmath5 $ ] is a maxwellian with non - constant ( non - isothermal ) wall temperature @xmath6 . based on new constructive coercivity estimates for both steady and dynamic cases , for @xmath7 and any fixed value of @xmath4 , we construct a unique solution @xmath8 to ( [ boltzmann ] ) and ( [ fdiffuse ] ) , continuous away from the grazing set and exponentially asymptotically stable .
this solution is a genuine non equilibrium stationary solution differing from a local equilibrium maxwellian . as an application of our results we establish the expansion @xmath9 and
we prove that , if the fourier law holds , the temperature contribution associated to @xmath10 must be linear , in the slab geometry .
this contradicts available numerical simulations , leading to the prediction of breakdown of the fourier law in the kinetic regime . |
the @xmath1discretization of this paper differs in an important way , both theoretically and practically , from the ` ordinary ' discretization studied in @xcite .
first , we point out what the two approaches have in common , then we underline their differences
. both discretizations must have an underlying compact simple or semisimple lie group @xmath0 of any rank @xmath11 and of any type .
the rank is equal to the number of variables involved in the process .
then we introduce new classes of multivariate special functions , the @xmath12 and @xmath13functions in @xcite , and the @xmath1functions ( [ e ] ) here .
the functions are orthogonal on a finite region @xmath14 of the @xmath15dimensional real euclidean space @xmath16 , as continuous functions as well as functions restricted to a fragment of a lattice @xmath17 .
the lattice can have any density but its symmetry is imposed by the underlying lie group .
thus in @xmath16 there are as many different lattices and special functions orthogonal on them as there are semisimple lie groups of rank @xmath18 .
the families of @xmath12 , @xmath13 , and @xmath1functions were recognized and named in @xcite .
they generalize the common cosine , sine , and exponential functions of one variable .
many of their properties are described in reviews @xcite .
the orbit function @xmath19 and @xmath20 are built using the finite reflection group @xmath21 , attached to each @xmath0 , and called the weyl group of @xmath0 .
the @xmath1functions are built using the even subgroup @xmath22 which is not a reflection group .
much less specific information is available about such groups in the literature .
the @xmath1functions are simpler , for the same @xmath0 , than the orbit functions of type @xmath19 or @xmath20 .
more precisely , an orbit function is a sum / difference of two @xmath1functions .
they have no prescribed behavior at the boundary @xmath23 of the region of their orthogonality @xmath24 , unlike the orbit functions which are either symmetric or antisymmetric with respect to their boundary @xmath25 .
region @xmath24 is a union of two adjacent copies of region @xmath14 .
discretizations of the functions over lattice fragments @xmath26 and @xmath27 , particularly their orthogonality over @xmath27 , require different specific values of a number of constants required for groups @xmath0 .
the purpose of the paper is to provide all of the information needed for the fourier analysis of @xmath15dimensional digital data in terms of their @xmath1function expansions , in the context of an admissible symmetry group @xmath0 .
we suppose that @xmath0 is one of the classical simple lie groups with the lie algebras of types @xmath28 , @xmath29 , @xmath30 , and @xmath31 . in section 2 , some standard properties of simple lie groups and/or simple lie algebras are recalled .
the properties of group @xmath32 not generally available elsewhere are important . in section 3
, we describe the lattice grids @xmath33 , where the digital data is provided .
the density of the grid is controlled by our choice of the integer @xmath6 . for any grid @xmath34
, there are only finitely many distinct @xmath1functions that are orthogonal on the grid .
also , the lowest set of such functions is described .
the functions are labeled by the grid of points @xmath7 . in section 4 ,
the properties of the @xmath1functions are described for each point of @xmath7 .
discrete @xmath1transforms are presented in section 5 .
concluding comments and remarks are found in section 6 .
consider the lie algebra of the compact simple lie group @xmath0 of rank @xmath18 , with the set of simple roots @xmath35 , spanning the euclidean space @xmath16 @xcite . by uniform and standard methods for @xmath0 of any type and rank ,
a number of related quantities and virtually all the properties of @xmath0 are determined from @xmath36 .
we make use of the following : * the highest root @xmath37 . here
the coefficients @xmath38 are known positive integers , also called the marks of @xmath0 . *
the coxeter number @xmath39 of @xmath0 . * the cartan matrix @xmath19 @xmath40 * the order @xmath41 of the center of @xmath0 , @xmath42 * the dual weight lattice , @xmath43 * the root lattice @xmath44 of @xmath0 , @xmath45 * the dual root lattice @xmath46 the properties of weyl groups and affine weyl groups can be found for example in @xcite .
the finite weyl group @xmath21 is generated by @xmath18 reflections @xmath47 , @xmath48 , in @xmath49-dimensional ` mirrors ' orthogonal to simple roots intersecting at the origin : @xmath50 the infinite affine weyl group @xmath51 is the semidirect product of the abelian group of translations @xmath52 and of the weyl group @xmath21 .
@xmath53 equivalently , @xmath51 is generated by reflections @xmath54 and reflection @xmath55 , where @xmath56 the fundamental region @xmath14 of @xmath51 is the convex hull of the points @xmath57 : @xmath58 since @xmath14 is a fundamental region of @xmath51 , we have : 1 . for any @xmath59 there
exists @xmath60 , @xmath61 and @xmath62 such that @xmath63 2 . if @xmath64 and @xmath65 , @xmath66 then @xmath67 , i.e. if there exist @xmath61 and @xmath62 such that @xmath68 , then @xmath69 3 .
consider a point @xmath70 , such that @xmath71 .
the isotropy group @xmath72 of the point @xmath73 is trivial , @xmath74 , if @xmath75 , where @xmath76 denotes the interior of @xmath14 , i.e. all @xmath77 , @xmath78 .
otherwise the group @xmath79 is finite and generated by such @xmath54 for which @xmath80 , @xmath78 .
elements of the weyl group @xmath21 are orthogonal linear transformations of the space @xmath16 . a subgroup of @xmath21 of the elements @xmath61 with determinant @xmath81 is called the even weyl group @xmath32 .
i.e. @xmath82 the subgroup @xmath32 can be viewed as the kernel of the homomorphism @xmath83 .
since @xmath84 and @xmath85 , group @xmath32 is a normal subgroup of @xmath21 and @xmath86 the infinite even affine weyl group @xmath87 is the semidirect product of the group of translations @xmath52 and of the even weyl group @xmath32 @xmath88 we choose some fixed @xmath89 and define the set @xmath24 by @xmath90 note that @xmath24 consists of two disjoint parts : the closed simplex @xmath14 and the open interior of the simplex @xmath91 . from this decomposition , we also obtain the formula for the volume of @xmath24 : @xmath92 in the following proposition , we show that @xmath24 is a fundamental region of the even affine weyl group @xmath87 .
[ tvrfund ] for the set @xmath24 , it holds that 1 . for any @xmath59 there
exists @xmath93 , @xmath94 and @xmath62 , such that @xmath95 2 . if @xmath96 and @xmath65 , @xmath97 then @xmath67 , i.e. if there exist @xmath94 and @xmath62 such that @xmath68 , then @xmath98 3 .
consider a point @xmath99
. if @xmath100 or @xmath101 , then the isotropy group @xmath102 is trivial , @xmath103 .
if @xmath104 , then it holds that @xmath105 1 .
suppose we have some @xmath59 .
it follows from ( [ fun1 ] ) that there exist @xmath60 , @xmath61 and @xmath62 such that @xmath106 .
if @xmath107 then we have found @xmath108 , @xmath94 and @xmath62 that satisfy ( [ fun1 ] ) .
suppose that @xmath109 and 1 .
since @xmath111 we have @xmath112 . taking into account that @xmath113 and @xmath114 we obtain @xmath115 and @xmath116 .
2 . @xmath117 .
we have from ( [ stab ] ) that the stabilizer @xmath118 is non - trivial and contains some @xmath54 , @xmath119 such that @xmath120 . if @xmath121 then we have @xmath122 and @xmath123 . if @xmath124 then we have @xmath125 and @xmath126 where @xmath127 .
suppose we have @xmath128 and @xmath94 , @xmath129 such that @xmath130 since @xmath24 consists of two disjoint parts @xmath14 and @xmath91 , we distinguish the following cases : 1 .
it follows immediately from ( [ fun2 ] ) that @xmath67 .
2 . @xmath131 .
consider @xmath132 such that @xmath133 and @xmath134 .
then @xmath135 and from ( [ fun2 ] ) we obtain @xmath136 , i.e. @xmath67 .
@xmath60 , @xmath137 .
consider @xmath138 , @xmath139 .
then @xmath140 and from ( [ fun2 ] ) we have that @xmath141 . since the stabilizer of the point @xmath139 is trivial , @xmath142 , we obtain @xmath143 .
we conclude that @xmath144 , which is contradictory to the assumption @xmath94 in ( [ funproof ] ) and thus , this case can not occur .
3 . if @xmath100 , then from ( [ stab ] ) we have that the stabilizer @xmath79 is trivial . since the stabilizers of the points @xmath73 and @xmath145 are conjugated , the stabilizer @xmath146 is also trivial . then since @xmath147 , we have @xmath148 .
+ we have from ( [ direct ] ) that for any @xmath66 there exist a unique @xmath61 and a unique shift @xmath149 such that @xmath150 .
define a homomorphism @xmath151 for @xmath152 by @xmath153 .
if @xmath154 then we have from ( [ stab ] ) that the stabilizer @xmath79 is non - trivial , finite and contains some @xmath54 , @xmath119 .
since @xmath155 , @xmath119 holds , we have @xmath156 .
then since @xmath157 , we conclude that @xmath158 .
if we have two elements @xmath160 such that @xmath161 , with @xmath62 , then for @xmath94 we have @xmath162 , i.e. we have a natural action of @xmath32 on the torus @xmath159 . for @xmath163 we denote the isotropy group and its order by @xmath164 we denote the orbit and its order by @xmath165 clearly we have @xmath166 1 . for any @xmath163 there
exists @xmath167 and @xmath94 such that @xmath168 2 . if @xmath169 and @xmath170 , @xmath94 then @xmath171 3 . if @xmath172 , i.e. @xmath173 , @xmath99 then @xmath174 1
follows directly from ( [ fun1e ] ) .
2 . follows directly from ( [ fun2e ] ) .
we have from ( [ directe ] ) that for any @xmath97 there exist a unique @xmath94 and a unique shift @xmath149 such that @xmath150 .
define a homomorphism @xmath175 for @xmath176 by @xmath177 . if @xmath178 then @xmath179 , i.e. we obtain @xmath180 and vice versa . thus , @xmath181 holds .
we also have @xmath182 the set of simple dual roots @xmath183 is a system of simple roots of some simple lie algebra .
the system @xmath184 also spans euclidean space @xmath16 .
the dual system @xmath184 determines : * the highest dual root @xmath185 . here
the coefficients @xmath186 are called the dual marks of @xmath0 . * the dual cartan matrix @xmath187 @xmath188 * the dual root lattice @xmath189 * the root lattice @xmath190 * the @xmath191-dual lattice @xmath192 dual affine weyl group @xmath193 is a semidirect product of the group of shifts @xmath44 and the weyl group @xmath21 @xmath194 equivalently , @xmath193 is generated by reflections @xmath54 and reflection @xmath195 , where @xmath196 the fundamental region @xmath197 of @xmath193 is the convex hull of the vertices @xmath198 : @xmath199 the dual even affine weyl group @xmath200 is the semidirect product of the group of translations @xmath44 , and of the even weyl group @xmath32 @xmath201 we choose some fixed @xmath89 and define the set @xmath202 by @xmath203 analogously to proposition [ tvrfund ]
, we obtain that @xmath202 is a fundamental region of the dual even affine weyl group @xmath200 . for the set @xmath202
, it holds that 1 . for any @xmath59
, there exists @xmath204 , @xmath94 and @xmath205 such that @xmath206 2 . if @xmath207 and @xmath65 , @xmath208 then @xmath67 , i.e. if there exist @xmath94 and @xmath205 such that @xmath209 then @xmath210 3 .
consider a point @xmath211 . if @xmath212 or @xmath213 , then the isotropy group @xmath214 is trivial , @xmath215 .
if @xmath216 then it holds @xmath217 where @xmath218
the grid @xmath8 is the finite fragment of the lattice @xmath219 which is found inside of @xmath24 .
suppose we have a fixed @xmath220 and consider the @xmath21-invariant group @xmath221 .
the group @xmath221 is finite with the order @xmath222 we define the grid @xmath8 as such cosets from @xmath221 which have a representative element in the fundamental domain @xmath2 : @xmath223 from the relation ( [ rfun1e ] ) , we have that @xmath224 the grid @xmath34 can be viewed as a union of two disjoint grids the grid @xmath225 and the reflection @xmath226 of its interior @xmath227 , @xmath228 we obtain from ( [ deffun ] ) that the set @xmath229 , or more precisely its representative points , can be identified as @xmath230 the reflection @xmath226 of its interior @xmath231 is given by @xmath232 the number of elements of @xmath34 could be obtained by combining results from @xcite .
[ nume ] let @xmath233 be the coxeter number .
then @xmath234 considering the equality @xmath235 , we obtain from the disjoint decomposition ( [ disfm ] ) that @xmath236 it was shown in proposition 3.5 in @xcite , that @xmath237 the numbers of points of the grid @xmath34 of lie algebras @xmath28 , @xmath29 , @xmath30 , @xmath31 are given by the following relations : 1 .
@xmath238 @xmath239 2 .
@xmath240 @xmath241 @xmath242 4 .
@xmath243latexmath:[\[\begin{aligned } for the case @xmath28 , we have from @xcite that @xmath245 and @xmath246 .
it can be verified directly that the formula @xmath247 satisfies ( [ intcases ] ) for all values of @xmath220 .
the result follows from ( [ soucet ] ) .
analogously we obtain formulas for algebras @xmath29 , @xmath30 and @xmath31 . using explicit formulas for @xmath248 from @xcite ,
the number @xmath249 of the five exceptional lie algebras can be obtained similarly from proposition [ nume ] .
[ ex1 ] for the lie algebra @xmath250 , we have coxeter number @xmath251 and @xmath252 .
consider for example @xmath253 .
for the order of the group @xmath254 , we have from ( [ grupa ] ) that @xmath255 and according to theorem [ numan ] we calculate @xmath256 the coset representants of @xmath254 and the fundamental domain @xmath24 are depicted in figure [ figc2 ] .
the points of @xmath257 are the weights that specify @xmath1functions belonging to the same pairwise orthogonal set .
further on , we consider @xmath1functions that are sampled on the points @xmath34 .
we consider the lowest possible set of such points .
the number of points of @xmath257 coincides with the number of points of @xmath34 .
the @xmath21-invariant group @xmath258 is isomorphic to the group @xmath221 and its order is given by @xmath259 define the grid @xmath257 as such cosets from @xmath258 with representative elements in @xmath260 @xmath261 the grid @xmath257 can be viewed as a union of two disjoint grids the grid @xmath262 and the reflection @xmath226 of its interior @xmath263 , @xmath264 we have from ( [ defdfun ] ) that the set @xmath265 , or more precisely its representative points , can be identified as @xmath266 the reflection @xmath226 of its interior is given by @xmath267
since the @xmath18tuple of dual marks @xmath268 is a certain permutation of the @xmath18tuple @xmath269 , we have from ( [ fm ] ) and ( [ lam ] ) that @xmath270 , and from ( [ rifm ] ) and ( [ rilam ] ) that @xmath271 . taking into account disjoint decompositions ( [ disfm ] ) and ( [ dislam ] )
, we conclude that @xmath272 if we have two elements @xmath273 such that @xmath274 , with @xmath205 then for @xmath94 we have @xmath275 , i.e. we have a natural action of @xmath32 on the quotient group @xmath276 . for @xmath277 we denote the order of the stabilizer @xmath278 1 . for any @xmath279 there exists @xmath280 and @xmath94 such that @xmath281 2 . if @xmath282 and @xmath283 , @xmath94 then @xmath284 3 . if @xmath285 , i.e. @xmath286 , @xmath287 then @xmath288 1
let @xmath279 be of the form @xmath289 , @xmath290 . from ( [ fun1ed ] )
it follows that there exist @xmath291 , @xmath94 and @xmath205 such that @xmath292 i.e. @xmath293 . from @xmath21invariance of @xmath294 we have that @xmath295 , the class @xmath296 is from @xmath257 and ( [ dfun1 ] ) holds .
2 . let @xmath297 be of the form @xmath298 , @xmath299 and @xmath300 .
suppose that @xmath301 then @xmath302 and it follows from ( [ fun2ed ] ) that @xmath303 .
we have from ( [ directed ] ) that for any @xmath208 there exist unique @xmath94 and unique shift @xmath304 such that @xmath305 .
define a homomorphism @xmath306 for @xmath307 by @xmath308 .
if @xmath309 then @xmath310 , i.e. we obtain @xmath311 and vice versa .
thus , @xmath312 holds .
we also have @xmath313 [ exdual ] for the lie algebra @xmath250 we have @xmath314 and according to ( [ fle ] ) we have @xmath315 the coset representants of @xmath316 , the dual fundamental domain @xmath202 and the grid @xmath317 are depicted in figure [ figc2d ] .
calculation procedure of @xmath320 for any @xmath321 and of @xmath322 , @xmath323 was derived in 3.7 of @xcite .
these calculation procedures use extended coxeter - dynkin diagrams ( @xmath324 ) of @xmath0 and their dual versions @xmath325 ( see e.g.@xcite ) . modifying these procedures by using the relations ( [ rfunstabe ] ) , ( [ stabe2 ] ) and ( [ rfunstab2 ] ) , ( [ stabed2 ] )
, we deduce a calculation procedure for @xmath318 , @xmath319 , defined by ( [ hxe ] ) , ( [ hla ] ) , for @xmath326 and @xmath327 .
consider a point @xmath328 . 1 .
if @xmath329 then @xmath330 .
2 . let @xmath331 $ ] be corresponding coordinates of @xmath332 from ( [ fm ] ) .
if @xmath333 are all non - zero , then @xmath330 .
if some of the coordinates @xmath331 $ ] are zero then consider such a subgraph @xmath334 of extended @xmath324 consisting only of those nodes @xmath335 for which @xmath336 .
the subgraph @xmath334 consists in general of several connected components @xmath337 .
each component @xmath337 is a ( non - extended ) @xmath324 of some compact simple lie group @xmath338 .
take corresponding orders of the weyl groups @xmath339 of @xmath338
. then it holds @xmath340 we proceed similarly to determine @xmath319 when considering a point @xmath341 . 1 .
if @xmath342 then @xmath343 .
2 . let @xmath344 $ ] be the corresponding coordinates of @xmath345 from ( [ lam ] ) .
if @xmath346 are all non - zero then @xmath343 .
if some of the coordinates @xmath344 $ ] are zero then consider such a subgraph @xmath347 of the extended @xmath325 consisting only of those nodes @xmath335 for which @xmath348 .
the subgraph @xmath347 consists in general of several connected components @xmath349 .
each component @xmath349 is a ( non - extended ) @xmath324 of some compact simple lie group @xmath350 .
take corresponding orders of the weyl groups @xmath351 of @xmath350
. then it holds @xmath352
the numbers @xmath318 , @xmath319 and @xmath249 , which were determined so far , are important for the properties of special functions , called @xmath1functions when they are sampled on @xmath34 .
a detailed review of the properties of @xmath1functions may be found in @xcite . in this section
the goal is to complete and make explicit the orthogonality properties of @xmath1functions @xcite .
we recall the definition of @xmath1functions and show that they can be labeled by the finite set @xmath257 when sampled on the grid @xmath34 .
consider @xmath354 and recall that ( normalized ) @xmath1functions can be defined as a mapping @xmath355 @xmath356 the following properties of @xmath1functions are crucial * symmetry with respect to @xmath94 @xmath357 * invariance with respect to @xmath358 @xmath359 we investigate values of @xmath1functions on the grid @xmath34 .
suppose we have a fixed @xmath220 and @xmath360 .
from ( [ einv ] ) it follows that we can consider @xmath361 as a function on classes @xmath221 . from ( [ rfun1e ] ) and ( [ esym ] ) it follows that we can consider @xmath361 only on the set @xmath362 .
we also have @xmath363 and thus we can consider the functions @xmath364 on @xmath34 parameterized by classes from @xmath279 .
moreover , from ( [ dfun1 ] ) and ( [ esymla ] ) it follows that _ we can consider @xmath1functions @xmath364 on @xmath34 parameterized by @xmath365 only_.
discrete orthogonality of @xmath1functions was discussed in general in @xcite . practical application of @xcite is not completely straightforward .
therefore we reformulate the basic facts and subsequently use them to make the discrete orthogonality over @xmath34 described in all detail .
basic orthogonality relations from @xcite are for any @xmath366 of the form : @xmath367 we define the scalar product of two functions @xmath368 by @xmath369 where the numbers @xmath370 are determined by ( [ epe ] ) . we show that @xmath257 , defined by ( [ dislam ] ) , is the lowest maximal set of pairwise orthogonal @xmath1functions . for @xmath371 it holds @xmath372 where @xmath41 , @xmath319 , @xmath364 were defined by ( [ center ] ) , ( [ hla ] ) , ( [ e ] ) respectively
, @xmath18 is the rank of @xmath0 .
the equality @xmath373 follows from ( [ rfun2e ] ) and ( [ wfm ] ) and @xmath353invariance of the expression @xmath374 .
then , using @xmath353invariance of @xmath221 and ( [ bdis ] ) , we have @xmath375 since @xmath371 we have from ( [ lrfun2 ] ) that @xmath376 the highest root @xmath377 and the highest dual root @xmath378 of @xmath250 are determined by the formulas @xmath379 the even weyl group of @xmath250 has four elements , @xmath380 , and we obtain for the determinant of the cartan matrix @xmath252 .
thus , decomposition of the grid @xmath381 is given by @xmath382 and the grid of weights @xmath383 is determined by @xmath384 the discrete orthogonality relations of @xmath1functions of @xmath250 which hold for any two functions @xmath364 , @xmath385 labeled by @xmath386 are of the form ( [ orthoe ] ) .
the coefficients @xmath370 , @xmath387 , which appear in ( [ orthoe ] ) , have according to 3.5 values @xmath388 , @xmath389 for @xmath390 , @xmath391 .
we represent each point @xmath392 and each weight @xmath393 by the coordinates @xmath394 $ ] and @xmath395 $ ] from relations ( [ c2me ] ) and ( [ c2le ] ) , respectively .
the values of the coefficients @xmath370 , @xmath387 for @xmath396 , @xmath397 are listed in table [ f4 ] . .the
coefficients @xmath370 and @xmath387 of @xmath250 . assuming @xmath398 , @xmath399 . [ cols="^,^",options="header " , ] analogously to ordinary fourier analysis , we define interpolating functions @xmath400 @xmath401 which are given in terms of expansion functions @xmath364 and expansion coefficients @xmath402 , whose values need to be determined .
these interpolating functions can also be understood as finite cut - offs of infinite expansions .
next we discretize the equations ( [ intc ] ) .
suppose we have some function @xmath403 sampled on the grid @xmath34 .
the interpolation of @xmath403 consists in finding the coefficients @xmath402 in the interpolating functions ( [ intc ] ) such that @xmath404 relations ( [ fle ] ) and ( [ orthoe ] ) allow the values @xmath405 with @xmath326 , @xmath327 to be viewed as elements of a non - singular square matrix .
this invertible matrix coincides with the matrix of the linear system ( [ intcs ] ) .
thus , the coefficients @xmath402 can be uniquely determined .
the formula for calculation of @xmath402 , which is also called discrete @xmath1transform , can be obtained by means of calculation of standard fourier coefficients : @xmath406 we also have the corresponding plancherel formula @xmath407
the @xmath1functions of the article have other undoubtedly useful properties that were not invoked here .
let us briefly mention six : * product of two @xmath1functions with the same underlying lie group and the same arguments @xmath408 but different subscripts , say @xmath409 and @xmath410 , decompose into the sum of @xmath1functions .
this is a powerful property that enables building of a set of recursion relations for constructing even larger @xmath1functions .
* points of the weight lattice @xmath294 of @xmath0 split into a few disjoint congruence classes .
it is convenient to specify a congruence class of @xmath411 by its congruence number @xcite .
there is a different linear function of @xmath412 for different lie group @xmath0 .
the weights of one @xmath21-orbit , hence also the weights of one @xmath353orbit , belong to the same congruence class . as a consequence , all summands @xmath413 in an @xmath1function
have the weights @xmath412 from the same congruence class .
hence , the @xmath1function has a well defined congruence number . during multiplication of @xmath1functions of the same @xmath0 ,
the congruence numbers of the @xmath1functions add up .
in particular , the decomposition of a product of two @xmath1functions must contain summands with the same congruence number .
the number of congruence classes of a @xmath0 is equal to the order of the center of the compact simple lie group @xmath0 . *
another undoubtedly useful property of orbit functions , which plays no role in this paper , is the fact that they are eigenfunctions of the laplace operator appropriate for the lie group , and that the eigenvalues are known in all cases @xcite . *
the @xmath1functions considered in the paper so far have the underlying simple group @xmath0 .
suppose now that it is semisimple but not simple , say @xmath414 , where @xmath415 and @xmath416 are simple .
then there are two options as to what to take for the even subgroup @xmath417 .
the simpler of the two is to define @xmath418 .
the @xmath1functions of @xmath419 are then products of the @xmath1functions of @xmath415 and @xmath416 , etc . , see @xcite .
such an option is trivial , for example when @xmath420 .
+ the option where @xmath417 is bigger , namely the full even subgroup of @xmath421 , is somewhat more interesting .
this options is still to be explored in the literature @xcite .
it is already non - trivial in the lowest case @xmath420 .
* the results presented in this paper are valid for any compact simple lie group @xmath0 , including the five exceptional cases .
explicit counting formulas for @xmath249 for these five cases , which were not provided here , can be straightforwardly put together using proposition [ nume ] and the appendix in @xcite . *
the present work raises the question : under which conditions converge series of the functions @xmath422 assigned to a function @xmath423 by the relations ( [ intc ] ) , ( [ ctrans ] ) .
work was supported by the natural sciences and engineering research council of canada and in part also by the mind research institute , santa ana , california .
jh is grateful for the postdoctoral fellowship and for the hospitality extended to him at the centre de recherches mathmatiques , universit de montral .
j. patera , _ compact simple lie groups and theirs @xmath12 , @xmath13 , and @xmath1transforms , _ sigma ( symmetry , integrability and geometry : methods and applications ) * 1 * ( 2005 ) 025 , 6 pages , math - ph/0512029 . r. v. moody and j. patera , _ orthogonality within the families of @xmath12 , @xmath13 , and @xmath1functions of any compact semisimple lie group , _ sigma ( symmetry , integrability and geometry : methods and applications ) * 2 * ( 2006 ) 076 , 14 pages , math - ph/0611020 . | three types of numerical data are provided for compact simple lie groups @xmath0 of classical types and of any rank .
this data is indispensable for fourier - like expansions of multidimensional digital data into finite series of @xmath1functions on the fundamental domain @xmath2 .
firstly , we determine the number @xmath3 of points in @xmath2 from the lattice @xmath4 , which is the refinement of the dual weight lattice @xmath5 of @xmath0 by a positive integer @xmath6 .
secondly , we find the lowest set @xmath7 of the weights , specifying the maximal set of @xmath1functions that are pairwise orthogonal on the point set @xmath8 .
finally , we describe an efficient algorithm for finding the number of conjugate points to every point of @xmath8 .
discrete @xmath1transform , together with its continuous interpolation , is presented in full generality .
@xmath9 centre de recherches mathmatiques , universit de montral , c. p. 6128
centre ville , montral , h3c3j7 , qubec , canada ; [email protected] + @xmath10 department of physics , faculty of nuclear sciences and physical engineering , czech technical university , behov 7 , 115 19 prague 1 , czech republic ; [email protected] |
spontaneous breaking of global symmetries in a gauge field theory sets in only when the gauge forces become strong . accordingly
, nonperturbative methods must be brought to bear , and the problem remains only partially understood .
approaches include lattice methods , semiclassical methods , and @xmath1 expansions .
in addition , a simple approach based in feynman graphs has been used with apparent success for many years .
it comes in several closely related forms , including schwinger dyson or gap equations , and `` most attractive channel '' analyses .
these share the common idea that the forces responsible for symmetry breaking can be computed in a loop expansion .
the lowest order contribution is typically one - particle exchange , with higher order contributions given by two particle irreducible graphs @xcite .
this can be applied to a most attractive channel force analysis or equivalently the kernel of a gap equation , or the cjt effective potential @xcite for composite operators . because the coupling must be relatively strong
, there is no obvious small parameter in the expansion and this approach has been met with justifiable skepticism . some evidence in favor of its utility , however ,
is provided by estimates of the second order contribution to the kernel @xcite . when the coupling is just strong enough to trigger the breaking , second order corrections are relatively small ( less than 20% ) . with this check in mind , and in the absence of clearly superior alternatives , gap equations have been widely used in recent years in the study of dynamical electroweak symmetry breaking @xcite . still , their validity remains open to question .
recent advances @xcite in nonperturbative methods for supersymmetric gauge theories provide an independent framework in which to test the validity of the gap equation approach .
the new methods lead to conclusions about spontaneous breaking of global symmetries in supersymmetric theories , which can be compared to gap equation results .
this letter reports the results of such a comparison , first for the case of @xmath2 supersymmetric yang - mills @xmath3 theory with no matter fields ( pure sqcd ) , and then for @xmath2 sqcd including matter fields .
the lagrangian for pure sqcd contains only gluon and gluino degrees of freedom ( in wess - zumino gauge after eliminating auxiliary fields ) : @xmath4 where the canonically normalized majorana spinor @xmath5 describes the gluino field , @xmath6 is the vacuum angle for the nonabelian gauge field @xmath7 , and @xmath8\,$ ] .
this theory has an anomalous @xmath9 global symmetry , @xmath10 , broken by instanton effects to @xmath11 .
various studies indicate the formation of a gluino condensate @xcite , which spontaneously breaks the discrete @xmath11 symmetry to @xmath12 .
nonperturbative holomorphic calculations derive this result starting from an exact wilsonian effective superpotential for @xmath13 , obtained @xcite from supersymmetric and global symmetry constraints , assuming the existence of a nonperturbative , supersymmetric regulator . successively integrating out massive flavors @xcite results in an effective superpotential for @xmath14 . differentiating with respect to the supersymmetric coupling @xmath15
then provides an exact formula @xcite for the gluino condensate , @xmath16 . here
, the subscript @xmath17 denotes a `` holomorphic '' normalization @xcite ( @xmath18 on the kinetic term ) for a weyl spinor , and @xmath19}$ ] is defined in terms of a holomorphic gauge coupling @xmath20 defined at an ultraviolet scale @xmath21 in the dimensional reduction scheme @xcite . for a canonical field normalization , rescaling @xmath22 requires the canonical coupling @xmath23 to satisfy @xcite @xmath24 , due to the transformation s anomalous jacobian @xcite .
we can accordingly express the holomorphic result as @xmath25 ( with canonically normalized , majorana spinors ) .
the one - loop @xmath26 function solution for @xmath23 diverges at the `` confinement '' scale @xmath27}$ ] .
gap equation techniques , too , indicate a @xmath11-breaking condensate , signalled by a dynamical gluino mass .
we write the gluino inverse propagator as @xmath28\ , i^{-1}$ ] , with the wave function factor @xmath29 defined for renormalized fields so that @xmath30 , and with the dynamical mass @xmath31 . in a gauge with gluon propagator
@xmath32/k^2 $ ] , the gap equation in euclidean space takes the form @xmath33 where @xmath34 and @xmath35 are respectively the larger and smaller of @xmath36 and @xmath37 .
( the angular integrations have been performed in four dimensions , after approximating @xmath38 ^ 2)$ ] by @xmath39 , and inserting an extra factor of @xmath39 in anticipation of the linearization discussed below . )
the structure of this equation is similar to that for the quark in ordinary qcd @xcite , differing only in that the group factor @xmath40 replaces @xmath41 .
we have neglected running of the coupling and of the gauge parameter @xcite , a crude approximation adequate to establish the existence of a symmetry - breaking solution in this theory and to set the stage for the presence of matter fields .
here , the approximation requires an ultraviolet cutoff on the integral .
when matter fields are included , an infrared fixed point will govern the transition and justify the neglect of running . a nontrivial solution to eq . ( [ gluinogap ] )
requires the coupling to exceed a critical value , near which the dynamical mass vanishes continuously @xcite .
this allows the critical coupling to be determined by analyzing the equation in the regime @xmath42 , where loop momenta @xmath43 dominate the integral and linearization in @xmath44 is a good approximation .
we can then also neglect corrections to the massless renormalization group formula @xmath45 , where the scaling exponent equals the gluino - field anomalous dimension @xmath46 we find solutions by @xcite inserting a scaling form @xmath47 . the scaling exponent @xmath48 , in leading approximation just ( half ) the gluino - mass anomalous dimension , obeys a @xmath49-independent equation to first order in @xmath50 : @xmath51 a solution requires the right hand side of this equation to be large enough for @xmath48 to become complex @xcite , corresponding to the mass anomalous dimension reaching unity .
the discrete global symmetry is predicted to break if the coupling exceeds the resulting critical minimum value , @xmath52 .
higher order corrections to the kernel of the gap equation may be small enough not to affect this conclusion qualitatively .
in nonsupersymmetric qcd , the next order corrections are less than 20% @xcite . in reality , the coupling runs ; this effect can be incorporated in a wkb - like approximation @xcite which predicts nonzero solutions at scales where the coupling exceeds the above critical value . since the theory is asymptotically free and presumably confining , that will occur at momenta of order the confinement scale .
function @xcite for @xmath14 : @xmath53 .
well before reaching the singularity in this @xmath26 function , the running coupling will reach the above @xmath54 .
the @xmath26 function s expansion parameter there is @xmath55 . ] the symmetry is thus predicted to break , and the confinement scale sets the size of @xmath44 via the full nonlinear gap equation .
inserting this result into the operator product formula for the gluino condensate leads to the estimate @xmath56 the last rough estimate , neglecting logarithmic factors , arises from dimensional analysis with a nonzero solution to the gap equation , identifying the confinement scale with @xmath57 in eq .
( [ gluinovev ] ) . for large @xmath58 , we may write this estimate in the form @xmath59 which has the same @xmath58 scaling ( with fixed @xmath60 ) as eq .
( [ gluinovev ] ) .
we note that a complete gap - equation prediction of symmetry breaking would require comparing the broken solution to the @xmath61 solution .
establishing the former to be the ground state would require settling currently unresolved issues of gauge dependence , resulting from truncating the effective action or the infinite set of coupled schwinger - dyson equations .
( even in qcd , gap equation techniques face the same question . ) assuming the broken solution to be the ground state ( with zero energy for unbroken supersymmetry ) , the gap equation and holomorphic approaches both predict that the discrete @xmath11 symmetry spontaneously breaks to @xmath12 via a gluino condensate .
including quark supermultiplets with @xmath0 flavors provides more scope for comparing the two approaches . again using component notation with auxiliary fields eliminated ,
we add @xmath0 quark flavors to the lagrangian , as dirac spinors in the @xmath3 fundamental representation ; their left- and right - handed components are respectively associated with scalar superpartners , @xmath62 and @xmath63 : @xmath64 where @xmath65 . in the limit of vanishing bare masses
@xmath66 , the global symmetries are : a discrete parity symmetry , @xmath67 ; chiral @xmath68 with quarks and squarks rotated together ; baryon number @xmath69 for quarks and squarks ; and the anomaly - free subgroup @xmath70 of the gluino / squark @xmath9 and the quark / squark axial @xmath71 .
although these symmetries forbid perturbative mass generation , nonperturbative condensates could break the symmetries , generating fermion and scalar dynamical masses and mixing the left- and right- scalars .
the scalar mixing would also require nonvanishing dynamical gluino mass , since @xmath70 combines gluino and axial rotations .
an important step in the nonperturbative study of supersymmetric theories was witten s observation @xcite that in a massive theory the difference between the number of bosonic and fermionic zero - energy states is conserved is an integer multiple of @xmath58 . ]
( the witten index ) , with a nonzero value ( @xmath58 for massive sqcd ) implying unbroken supersymmetry . in that case ,
supersymmetric ward identities for @xmath72 leave fermion condensates proportional to the bare mass .
specifically , for the gluino condensate the konishi anomaly @xcite gives @xmath73 .
for the quark condensate @xcite , each flavor @xmath74 satisfies @xmath75 . the massless theory
thus forbids fermion condensates , assuming a well - defined , nonvanishing index at @xmath76 , and assuming squark bilinear expectations are not too singular when @xmath66 .
the holomorphic effective action approach @xcite provides a more complete description of sqcd with @xmath0 flavors .
assuming the existence of a nonperturbative , supersymmetric regulator , it leads to a self - consistent picture for various values of @xmath0 .
many physically distinct but ( if @xmath77 ) degenerate vacua form a moduli space , on which generically the chiral symmetry breaks , but at special points ( when expectation values vanish ) is preserved .
+ @xmath78 for @xmath79 , there is no asymptotic freedom , and the quarks , squarks , gluons and gluinos become noninteracting at large distances .
+ @xmath78 for @xmath80 , the theory is asymptotically free and the coupling runs to a `` superconformal '' infrared fixed point . there is no confinement in this `` nonabelian coulomb phase '' , and the global symmetries remain unbroken at the origin of moduli space ( where the squark vacuum expectation values vanish ) .
the spectrum is described by interacting , massless quarks , squarks , gluons , and gluinos .
+ @xmath78 for @xmath81 , there is no confinement , but the gauge theory is strongly interacting in the infrared .
the spectrum is best described in terms of massless , composite mesons and baryons of a local , infrared - free , dual gauge theory . at the origin of moduli space the theory leaves unbroken all the original global symmetries .
+ @xmath78 for @xmath82 or @xmath83 , the theory confines and the spectrum consists of massless composite particles corresponding to meson and baryon fields composed of the original matter fields . for @xmath82 , either @xmath69 or chiral @xmath68 must break . for @xmath83
, there can be confinement without chiral symmetry breaking .
+ @xmath78 for @xmath84 , a nonzero superpotential @xcite lifts the degeneracy and fixes the ground state at arbitrarily large scalar expectation values ( for @xmath66 ) .
+ an important feature of these results is that for @xmath85 , the global symmetries associated with massless quarks , squarks and gluinos can all remain unbroken .
we next use the gap equation framework to study the sqcd theory with matter .
a previous investigation @xcite concentrated on the quark and squark dynamical masses , neglecting the gluino condensate by assuming the matter fields to occupy a high dimensional representation . here
we will retain all condensates , to obtain a set of coupled gap equations for the quarks , gluinos , and squarks .
the most attractive interactions occur in channels that preserve local @xmath3 and parity , so we will not consider color - breaking condensates , or differing dynamical masses for the scalars @xmath62 and @xmath86 ( eq .
[ sqcd - l ] ) .
we will also exclude nonzero vacuum values for the squark fields .
supersymmetry requires equal bare masses for the quarks and squarks , which we take to vanish ; but the dynamical masses can differ , since the component notation does not maintain manifest supersymmetry in off - shell green functions . for small enough coupling , the running will be determined by the ( scheme - independent ) two - loop @xmath26 function @xcite @xmath87 which displays an infrared fixed point when @xmath88 .
it occurs at @xmath89 a small value for @xmath0 sufficiently close to @xmath90 .
the coupling approaches this fixed point at scales below some intrinsic scale @xmath91 governing the solution to eq .
( [ susybeta ] ) .
if the fixed point coupling exceeds a critical value , then the gap equations will exhibit nontrivial solutions .
since the dynamical mass vanishes continuously there , the transition will be governed by the fixed point , and the running of the coupling may be neglected to first approximation .
furthermore , the gap equations may be linearized in the neighborhood of the transition .
the gluino and quark gap equations are shown in fig . 1 .
to lowest order in the kernel and in the linearized approximation relevant to the study of the transition and the determination of the critical coupling , they each receive a contribution only from the first graph ( gluon emission and reabsorption ) .
the gluino gap equation is simply the linearized version of eq .
( [ gluinogap ] ) , and the corresponding quark equation takes the form @xcite @xmath92 where @xmath50 is now the value of the coupling at the infrared fixed point . for both , the intrinsic scale @xmath91 is effectively an ultraviolet cutoff in this approximation : the functions @xmath93 steepen there from near - critical behavior @xmath94 , to asymptotic behavior @xmath95 .
the gauge parameter is also in general a function of momentum , governed by a renormalization group equation .
when @xmath50 approaches its infrared fixed point , the gauge parameter has one as well .
we denote this fixed - point value simply by @xmath49 , and in this approximation take it outside the integral .
the wave function factors @xmath96 are computed in the massless theory , dropping corrections subleading in mass .
they are defined to be unity at the renormalization scale @xmath97 . for the scalars
, we obtain a gap equation for the diagonal mass @xmath98 , appearing in both @xmath99 and @xmath100 : @xmath101 \bigr\ } \biggr ] \nonumber\end{aligned}\ ] ] where the wave function renormalization constant @xmath102 implements @xmath103 , and @xmath104 represents the gluon wave function factor . and @xmath49 are in general functions of @xmath105 . as in the fermion equations ,
this dependence can be neglected if the integrals are dominated by an infrared fixed point . ]
the dynamical mixing mass @xmath106 in @xmath107 satisfies @xmath108\ .\end{aligned}\ ] ] in eqs .
( [ squarkdgap ] ) and ( [ squarkxgap ] ) we have dropped some contributions from diagrams with higher powers of mass insertions , as well as higher order contributions from diagonalizing the scalar propagator .
supersymmetry ensures cancellation of the quadratic divergences in the diagonal mass equation ( [ squarkdgap ] ) .
after that cancellation , both equations can be fully linearized for momenta @xmath37 large compared to @xmath109 .
then up to mass - suppressed corrections , @xmath110 is fixed by eqn .
( [ squarkdgap ] ) , while @xmath45 and @xmath111 , where @xmath112 here @xmath113 , and @xmath50 and @xmath49 are the fixed point values described above .
since the critical couplings depend on the wave function factors , the gluino critical coupling will differ from its value in the matter - free theory , even though in both cases gluon emission and reabsorption is the only leading contribution to the gap equation .
we focus initially on the quark and gluino equations . as described for the matterless theory ,
substituting the wave function factors of eq .
( [ anomdimsqcd ] ) into the gap equations , and setting the mass anomalous dimensions to unity , leads to critical couplings independent of @xmath49 : @xmath114 the choice of @xmath0 determines whether the infrared coupling @xmath115 in eq .
( [ fixedpoint ] ) achieves these critical values . as @xmath0
is reduced from @xmath90 , @xmath115 increases from zero and first exceeds the quark critical coupling , @xmath116 , when @xmath117 . for this value of @xmath118
, @xmath115 does not yet reach the gluino critical coupling , leaving quark condensation to play the primary role here . assuming validity of the approximations above
, we can now outline the phase structure of sqcd as described by gap equations : + @xmath78 for @xmath79 there is no asymptotic freedom , in agreement with the holomorphic description .
+ @xmath78 for @xmath119 there is an infrared fixed point at which the two - loop @xmath26 function vanishes .
for @xmath0 close enough to @xmath90 , the fixed point coupling is small , keeping the coupling well below the critical values of eq .
( [ critcouplsqcd ] ) .
for any @xmath0 above @xmath120 , the infrared coupling remains below both critical couplings , leaving the quark and gluino symmetries unbroken .
confinement does not set in and the theory remains in the nonabelian coulomb phase , in agreement with the holomorphic prediction .
+ @xmath78 for @xmath121 , the infrared coupling ( in the two - loop approximation ) exceeds the quark critical coupling .
if higher order effects may be neglected , the gap equation then indicates that quarks develop a dynamical mass which vanishes continuously at the transition .
the quarks decouple from the @xmath26 function at this mass scale , eliminating the fixed point and causing @xmath23 to run to the gluinos ( larger ) critical coupling .
gluino and scalar dynamical masses are then generated at essentially the same scale as the quark mass .
below these mass scales , decoupling allows @xmath23 to continue running to confining values .
the global chiral and @xmath70 symmetries are thus predicted to break in this @xmath0 range .
the supersymmetric ward identity then requires that supersymmetry breaks as well .
+ holomorphic methods , by contrast , predict the nonabelian coulomb phase to remain down to @xmath122 , and allow preservation of the global symmetries without confinement even down to @xmath83 .
these patterns also control the scalar gap equation solutions , which depend upon the fermion solutions .
the inhomogeneous scalar equations do not admit vanishing solutions when the quark and gluino dynamical masses are nonzero @xcite . in that case
, the scalar solutions should be of comparable scales .
the connection between scalar and fermion mass solutions can be understood from the global symmetries : a nonzero off - diagonal mass @xmath106 breaks both @xmath70 and @xmath68 , and thus requires corresponding gluino and quark masses .
the diagonal mass @xmath98 breaks no chiral or @xmath123 symmetries , but is not forced to vanish unless the quark mass does and supersymmetry is unbroken . checking that these symmetry breaking solutions of the gap equation describe
the ground state would require showing that the resulting vacuum energy is lower than that of the chirally symmetric solution .
( for this to be the case the symmetric solution , as well as the broken one , would need to break supersymmetry . ) for sqcd with matter , suggestive calculations @xcite do indicate lower energy for the symmetry breaking vacuum , although these are gauge dependent and neglect squark and/or gluino dynamical masses .
in pure sqcd without matter fields , various approaches indicate that the discrete global symmetry @xmath11 breaks spontaneously to @xmath12 through the formation of a gluino condensate . the theory confines and
supersymmetry presumably remains unbroken . in the gap equation
, the lowest order kernel , single gluon exchange , provides an attractive force capable of condensate formation . whether the broken solution represents the ground state of the theory
, however , remains unclear . assuming this to be the case , the gap equation and holomorphic effective action techniques lead to the same conclusion . for sqcd with matter multiplets ,
gluon exchange alone again gives the leading contribution to the linearized gap equation governing the transition . at this order ,
the gap equation ( with gauge coupling governed by the two - loop @xmath26 function ) indicates that spontaneous breaking of the global symmetries occurs for @xmath0 below a critical value near @xmath120 .
this transitional value corresponds to the infrared fixed point coupling exceeding the critical strength necessary for quark condensation . if we instead determine the fixed point coupling from the three - loop @xmath26 function ( in the dr scheme ) @xcite , the corresponding critical @xmath124 , only an 8% shift in this renormalization scheme .
examining the size of higher order terms in the kernel of the quark gap equation , corresponding to the mass anomalous dimension at two loops , would give another check on the validity of these approximations .
in nonsupersymmetric theories , such kernel corrections are of order 20% @xcite . with these caveats , for @xmath0 just below the critical value the quark mass is nonzero but small , justifying the linearization . at scales below their predicted mass
the quarks decouple , eliminating the infrared fixed point so that the coupling increases to the gluino critical value at essentially the same scale .
the gluinos , too , thus condense at the quark mass scale .
squark masses are induced by the fermion masses , and are also of the same scale .
the supersymmetric ward identity then predicts that supersymmetry also breaks in this regime .
an effective theory with goldstone bosons and goldstinos emerges at low energies .
all this is in contrast with the picture emerging from holomorphic effective action techniques , where the nonabelian coulomb phase is argued to persist down to @xmath122 and the global symmetries to remain unbroken ( at the origin of moduli space ) down to @xmath83 . of course , as already stressed , gap equation solutions in general
have not been shown to correspond to ground states .
( supersymmetric constraints on ground state energies may be important in this connection . ) settling the question involves still - unresolved issues of gauge invariance and interpretation of the effective potential .
furthermore , truncating kernels at one loop , and the @xmath26 function at two loops , seems quantitatively less reliable than in nonsupersymmetric theories .
the kernel truncation , together with the use of component notation ( wess - zumino gauge ) , may also in effect explicitly break supersymmetry
. the latter possibility can be checked by analyzing a set of gap equations in the manifestly supersymmetric superspace formalism @xcite .
a recent study of this problem @xcite indicates that symmetry - breaking solutions could exist , although infrared divergences in the formalism have so far obscured the analysis .
the possible discrepancy between gap equation analyses and results based on holomorphic effective action techniques and index theorems is noteworthy and invites further study .
efforts to reconcile them may help deepen our insight into the behavior of strongly coupled gauge theories .
the authors are grateful to nick evans , stephen hsu and myckola schwetz for their assistance in the early stages of this work .
we thank them and andrew cohen , noriaki kitazawa , aneesh manohar , gregory moore , pierre ramond and john terning for helpful discussions .
this work was supported in part under united states department of energy doe - hep grant de - fg02 - 92er-4074 . a.n .
also acknowledges the support of schweizerischer nationalfonds . for a review ,
see t. appelquist , yale preprint yctp
p2391 , published in _ proc .
mexican school of particles and fields , 4th , oaxtepec , mexico , dec .
1990 _ , ed .
lucio and a. zepeda ( world scientific , 1992 ) .
g. veneziano and s. yankielowicz , phys .
b113 , 231 ( 1982 ) ; + t.r .
taylor , g. veneziano and s. yankielowicz , nucl .
b218 , 493 ( 1983 ) ; + m.a .
shifman and a.i .
vainshtein , nucl .
b296 , 445 ( 1988 ) ; + d. amati , k. konishi , y. meurice , g.c .
rossi and g. veneziano , phys .
162 , 169 ( 1988 ) . | we compare gap equation predictions for the spontaneous breaking of global symmetries in supersymmetric yang - mills theory to nonperturbative results from holomorphic effective action techniques . in the theory without matter fields , both approaches describe the formation of a gluino condensate . with @xmath0 flavors of quark and squark fields , and with @xmath0 below a certain critical value , the coupled gap equations have a solution for quark and gluino condensate formation , corresponding to breaking of global symmetries and of supersymmetry .
this appears to disagree with the newer nonperturbative techniques , but the reliability of gap equations in this context and whether the solution represents the ground state remain unclear
. 16.5 cm 22.5 cm 0.2 cm -1.75 cm -2.2 cm 0.8 cm ( 0,0)(0,0 ) ( 365,245)hep - th/9709177 ( 365,230)yctp p1297 ( 365,215)wustl hep9755 |
in recent years , the question of photometric identification of supernova candidates has emerged as one of the crucial issues to be resolved before the advent of large - space supernova cosmology experiments , both ground - based ( _ e.g. _ , the large synoptic survey telescope [ lsst ] , the dark energy survey [ des ] , the panoramic survey telescope and rapid response system [ pan - starrs ] ) , and space - based ( _ e.g. _ , the _ joint dark energy mission [ jdem ] _ ) .
there are a number of reasons for this .
first , although there have been some interesting developments in the possible uses of supernova other than type ia for cosmology @xcite , type ia supernovae ( snia ) remain the staple of experimental cosmology .
second , sne ia are most reliably identified using spectroscopy due to the presence of a characteristic siii line at 6150 in the supernova rest frame . however , future large ground - based surveys are expected to collect thousands of supernova candidates , making a spectroscopic follow - up of each candidate all but unrealistic .
the identification of supernova candidates ( with possible spectroscopic follow - up for a select sample ) based on broadband photometry remains the only feasible alternative .
there have been a number of methods proposed to identify supernovae using broadband photometry that can be divided into three broad categories .
one includes methods that rely on color - color or color - magnitude diagrams @xcite .
it is also possible to fit supernova data to models , and select the best fit ( using , for example , a @xmath0 ) , which can be used to represent the supernova type @xcite . finally , the third category involves recently developed techniques based on a probabilistic ( bayesian ) approach to the problem @xcite .
the method proposed in this work , although closer in spirit to the second category , has a number of advantages over both . the existing techniques , while adequate in many cases , have a number of serious shortcomings .
for example , supernova identification schemes based on color - color and color - magnitude diagrams involve comparing the colors and/or magnitudes of a given supernova candidate with what is predicted by various supernova models .
this is an intuitive approach , allowing one to visually judge the goodness of fit of the data to the models ; however , it is difficult to account for all statistical and systematic uncertainties in a single step . a class of techniques that could be generally described as `` @xmath0-based '' simply find the best fit for a given supernova candidate s light curves to a supernova model .
this is also an intuitive and often completely reasonable approach , which nevertheless suffers the following disadvantages : 1 .
[ item:1 ] `` this object is not a supernova of any kind '' is not a well - defined hypothesis in this as in any other frequentist approach @xcite .
conversely , if the data happen to have large uncertainties , there is the possibility that a number of supernova models will be good fits to the data .
there is no formalism to compute not only the probability that a given fit is good , but also that it is bad . in other words ,
what one is interested in is the posterior probability , the probability that a given hypothesis is true given the data . calculating this probability
requires that the probability that this hypothesis is false be also known .
3 . using a @xmath0-based technique only gives the information about the best fit for a given set of data to a model , while any information about worse fits is lost .
the best fit will not necessarily reflect the true properties of the supernova .
4 . in cases where one would like to use a tail probability for accepting or rejecting given supernova candidates ( _ e.g.
_ , as sne ia ) , the probability of falsely rejecting the null hypothesis ( the so called type i error rate ) can be shown to be severely underestimated ( see @xcite and references therein ) .
bayesian classification schemes address many of the problems of the above - mentioned methods .
however , existing bayesian - based supernova typing methods have a serious drawback : they require the knowledge of the complete set of objects that a supernova candidate might conceivably be @xcite .
that is , they assume that a supernova candidate can only be one of a finite set of supernova types .
however , even with the current small high - redshift sn sample ( obtained almost exclusively with the hubble space telescope ) one occasionally finds supernova candidates with surprising new properties that do not seem to conform to any known models @xcite .
problems with assuming a finite set of supernova - like objects are further addressed in section [ sec : danger ] . in our work
, we introduce a likelihood ratio ( a bayes factor ) that is capable of discriminating between sne ia and _ anything else _ based on broadband photometric measurements .
the most important feature of this technique is that it is independent of the knowledge of the complete set of objects that a supernova candidate might conceivably be .
another advantage is that , as with all bayesian - based techniques , this method allows one to include all possible statistical and systematic uncertainties in a single step .
finally , the bayes factor is formulated in terms of color and thus does not require that one make any assumptions about the absolute magnitudes of the supernova candidates in the broadband filters used .
of course it is often desirable to include magnitudes in the formalism ; however , not only does it require making assumptions about the distribution of magnitudes for various known supernova types , but also it places a hard upper limit on the intrinsic magnitudes of objects that have yet to be observed . but more importantly
, `` anomalous '' ( non - supernova ) objects can be defined in a far more mathematically elegant and computationally manageable way using color alone .
the bayes factor is defined as @xmath1 where @xmath2 is the probability of obtaining the observed photometry ( colors ) from a sn ia , and @xmath3 is the probability of obtaining the data for any other object ( which could be a non - sn ia or any other object capable of mimicking an sn ia signal ) .
both probabilities take into account the _ relative _ distribution of light among the broadband filters used for the measurements . in general , no specific set of models ( or templates ) for non - sne ia is required for the calculation of the denominator . on a more technical note
, it is worthwhile to point out that bayes factors are normally used for deciding on the best of two hypotheses .
this allows one to easily set thresholds on the bayes factor in terms of the so - called type i and type ii error rates in the same way as thresholds are set on the likelihood ratio in @xcite .
also , although the main focus of this work is to describe a method that can identify sne ia , the bayes factor can be easily cast in terms of a posterior probability that a candidate is a type @xmath4 supernova , where @xmath4 could be ibc , ii - p , iin , _
etc._. , a hypothesis @xmath5 , and its alternative @xmath6 .
the bayes factor can be defined as @xmath7 the posterior probability that the alternative hypothesis is true for the data can then be written in terms of @xmath8 provided that one knows the priors for @xmath5 and @xmath6 , denoted by @xmath9 and @xmath10 , respectively : @xmath11 see @xcite for details .
although there are historical reasons why the bayes factor is formulated in this way , it is also convenient when setting thresholds for the error rates because often the errors rates are at least somewhat determined by the information contained in the priors on @xmath5 and @xmath6 . ]
this paper is organized as follows . in section [ section :
math ] we derive an expression for @xmath8 for a number of different cases .
we describe the performance of the method in section [ section : performance ] . section [ section : disc ] presents a discussion of the results .
the bayes factor , @xmath8 , introduced above , is defined as the probability of obtaining the photometric measurements assuming that the supernova candidate is a type ia over the probability that it is anything else . in practice ,
the probability that a candidate is an sn ia is the probability that the colors are consistent with what is expected for an sn ia using some prior knowledge about the behavior of type ia s . in our first formulation of the bayes factor , if the candidate is not in fact an sn ia , then the distribution of light in the broadband filters used can be arbitrary .
however , one could argue that much of the background for sne ia will be supernovae of other types whose behavior is relatively well - known
. however , the unprecedented scale of the future supernova surveys makes it highly likely that many new types of transient objects will be discovered .
also , little is known about the _ rates _ of non - type ia supernovae , especially at very high redshifts , making it difficult to predict the behavior of the background at those redshifts .
we begin with a general overview of the calculation of @xmath8 . for simplicity
, we assume that there are only two broadband filters , and that there is a single measurement of the supernova candidate s flux in each .
suppose that the flux is measured in photon counts , and that @xmath12 counts are measured in the first filter , and @xmath13 , in the second .
further suppose that there exists a model ( a template ) for the behavior of sne ia in these filters , and that the model predicts that some mean fraction of photons , @xmath14 , must end up in the first filter , and 1-@xmath14 , in the second .
the numerator of @xmath8 , @xmath2 , is essentially the probability that the measurement is consistent with this model .
assuming poisson statistics for the photon distributions , it can be easily shown that @xmath2 takes the form of a standard binomial distribution : @xmath15 for the calculation of the denominator of @xmath8 , @xmath3 , we do not make any _ a priori _ assumptions about the fraction of light that will end up in either filter .
the bayesian framework of the calculation allows one to circumvent this difficulty by marginalizing , or integrating over , all possible fractions .
mathematically , @xmath16 in reality , the calculation becomes rather more complicated . to begin with
, the measured flux will most likely be better described using gaussian , rather than poisson , statistics .
we must also allow for the possibility of multiple measurements and more than 2 filters .
in the next section we will make the the calculation more explicit and account for all of these factors .
before we plunge into the full derivation of @xmath8 for the case of gaussian statistics and multiple measurements and filters , we take a closer look at the simple case of a single measurement of a supernova candidate in just two filters , assuming that the photon count fluctuations in the filters are poisson . recall that we assume that @xmath12 counts are measured by the first filter and @xmath13 by the second ; and that we have a model that predicts a certain fraction of photons , @xmath14 , for the first filter , and @xmath17 , for the second .
following a similar derivation in @xcite , let us now introduce two variables , @xmath18 and @xmath19 , such that the mean number of photons in the first filter is given by @xmath20 , and the mean number of photons in the second filter is given by @xmath21 .
variable @xmath19 ranges from 0 to @xmath22 , and can be thought of as the mean number of photons that are counted in both filters for a given measurement .
variable @xmath18 ranges from 0 to 1 , and can be thought of as the probability that the photons will end up in the first filter as opposed to the second .
an analogy would be collecting balls into two receptor bins with different volumes : in this case , @xmath19 would be the mean number of balls that will enter both bins , and @xmath18 is the relative `` acceptance '' of one bin .
the introduction of these variables allows us to expand the bayes factor , eqn.[eqn : general ] , in terms of @xmath18 and @xmath19 : @xmath23 here , the first term in the numerator , @xmath24 , is the likelihood of obtaining the measurement given that the mean number of photons was measured to be @xmath25 in the first filter , and @xmath26 in the second .
likewise , the first term in the denominator , @xmath27 , is the likelihood of obtaining the measurement given that the mean number of photons was measured to be @xmath20 in the first filter , and @xmath21 in the second .
note that the numerator is not a function of @xmath18 because @xmath18 is single valued in the numerator , @xmath28 .
if the photon distribution is governed by poisson statistics , then : @xmath29 and @xmath30 the terms @xmath31 in the numerator and @xmath32 in the denominator of eqn.[eqn : general_expanded ] , are prior probabilities containing information regarding the expected distribution of light in the two filters for an sn ia and anything else , respectively . defining @xmath33 and @xmath34 as the minimum and maximum bounds for @xmath19 and assuming each value for @xmath19 in between these bounds is equally probable , we have : @xmath35 note that the range of @xmath19 will always be assumed to be @xmath36 $ ] , although the upper and lower bounds will initially be set to @xmath34 and @xmath33 , respectively . _
i.e. _ , to avoid priors whose probabilities approach 0 .
] however , if the candidate is not an sn ia , we do not make any assumptions about what to expect , and so : @xmath37 as the upper ( @xmath38 ) and lower ( @xmath39 ) bounds of @xmath18 are 1 and 0 , respectively .
note that @xmath31 and @xmath32 are _ improper priors _ ( in other words , they assume probability density functions that are flat and are integrated from zero to infinity ) .
this is not a major issue for our calculation because the priors happen to cancel .
however , in general the use of improper priors must be treated with caution @xcite .
it is therefore important to check that @xmath8 does indeed behave properly ; this will be addressed further in section[section : results ] . with these priors , eqn .
[ eqn : general_expanded ] becomes : @xmath40 in the calculation of eqn.[eqn : intermediate1 ] @xmath19 is effectively unconstrained , leaving the supernova candidate s magnitude free to take on any value .
that is , as we are only concerned with the relative fraction of photons in each filter ( _ i.e. _ , color ) , we need not make any assumptions about the behavior of @xmath19 .
we would now like to derive an equation analogous to eqn .
[ eqn : intermediate1 ] , but for the case of gaussian statistics .
let us suppose that instead of measuring @xmath12 photons in the first filter and @xmath13 in the second , we now measure a flux @xmath41 in the first filter with an error @xmath42 , and a flux @xmath43 in the second filter with an error @xmath44 .
as before , we parametrize the mean ( or `` true '' ) fluxes in the two filters as @xmath20 and @xmath21 , and expand @xmath8 in terms of @xmath18 and @xmath19 , leading to eqn .
[ eqn : general_expanded ] .
we then simply replace the poisson distributions with gaussian ones using the usual notation for a gaussian distribution , @xmath45 .
equation[eqn : general_expanded ] becomes @xmath46 the integration over @xmath19 in eqn .
[ eqn : gaussian ] can be reduced further leading to the appearance of the gauss error function .
however , the integration over @xmath18 in the denominator can only be done numerically .
we now make eqn .
[ eqn : gaussian ] even more realistic by considering multiple measurements ( say , @xmath47 ) and an arbitrary number of filters ( say , @xmath48 ) . using the formalism we have developed above
, we will assume that for the @xmath49 measurement the fraction of light in the @xmath50 filter is @xmath51 , and the total light distributed between all the filters and all the measurements is given by @xmath19 .
therefore , the hypothesized flux in the @xmath50 filter and @xmath49 measurement is given by @xmath52 . again , if the supernova is assumed to be a type ia , then we have a model that describes the fraction of light in each of the filters for each of the measurements must be .
the model must take into account the many possible observational parameters that characterize an sn ia .
for example , it is known that sne ia have a variety of possible `` stretch '' values , which parametrize the width of their light curves @xcite .
following the approach used in @xcite , we represent the type ia supernova parameters by @xmath53 , defined as @xmath54 where @xmath55 is the stretch parameter ; @xmath56 and @xmath57 parametrize the effect of interstellar dust extinction using the cardelli - clayton - mathis ( ccm ) parametrization @xcite ; @xmath58 accounts for the difference in the time of maximum of the model and the data ; and @xmath59 is the redshift . in other words , for type ia s @xmath51 will become @xmath60 .
the exact assumptions about the distribution of these parameters will be discussed below when the prior probabilities for all of the @xmath53 parameters will be stated explicitly . since , in general , the exact values of each of these parameters for a given candidate are unknown , they must be marginalized .
if the @xmath61 measurement in filter @xmath62 of the flux is @xmath63 with error @xmath64 , the multi - measurement , multi - filter analog of eqn .
[ eqn : general_expanded ] is : @xmath65 where @xmath66 , and @xmath67 indicates an integration over the multi - dimensional parameter space where @xmath68 .
the denominator is not parametrized by @xmath69 as it is not known what parameters are relevant for what we define as `` anything other than sne ia '' .
therefore , every possible distribution of light in the filters is given an equal chance .
we now address each term in eqn .
[ eqn : terms ] in turn .
@xmath70 + and @xmath71 are the likelihoods of obtaining a set of fluxes , @xmath72 , with uncertainties @xmath73 , for a number of measurements and filters , given that the mean number of photons are measured to be @xmath74 and @xmath75 , respectively .
assuming each measurement and every filter are independent , @xmath76 and @xmath77 note that , in general , the measured flux ( @xmath63 ) and the hypothesized flux ( @xmath78 ) have different subscripts ( which indicate the measurement number ) .
this is done to emphasize the fact that it is unknown where the time of maximum of our measured light curve is relative to that of the model .
this uncertainty is taken into account in one of the @xmath53 parameters , @xmath58 .
the terms @xmath79 and @xmath80 in eqn .
[ eqn : terms ] are prior probabilities . in particular , @xmath79 contains the prior knowledge about the parameters @xmath53 that describe an sn ia . for @xmath80 ,
@xmath81 is not constrained .
likewise , parameter @xmath19 is not constrained in any way for either a type ia or a non - ia prior .
it is therefore marginalized . integrating over @xmath19 means integrating over the total light in all the filters and all the measurements for a given candidate _
i.e. _ , integrating over the observed magnitude for this measurement .
furthermore , in marginalizing @xmath19 , there is an implicit assumption about the prior distribution of @xmath19 namely , that it is flat .
this assumption allows us to formulate the probabilities purely in terms of color .
allowing the total light to vary measurement - by - measurement with a flat prior is arguably the lightest possible assumption one can make regarding the magnitude .
explicitly , the priors @xmath82 and @xmath83 become : @xmath84 since the only constraint here is that all the light fractions in different filters add up to one ; and @xmath85 where @xmath86 is a the prior probability of @xmath53 . the priors on @xmath53 are defined similarly to those in @xcite and are briefly summarized below .
the stretch parameter @xmath55 follows a gaussian distribution with a mean of @xmath87 and a width of @xmath88 ( these values are extracted from @xcite ) .
the ccm parameters @xmath56 and @xmath57 can assume two sets of values with equal probabilities : @xmath89 ( no extinction ) and @xmath90 ( moderate extinction ) .
the prior probability for each choice of @xmath56 and @xmath57 is therefore @xmath91 .
the parameter accounting for the difference between the time of maximum of the data and the model , @xmath58 , has a flat prior .
the measured light curve is shifted relative to a template in one day increments 1000 times and each shift is assigned an equal probability @xmath92 .
means that a flat prior is assigned to @xmath58 .
finally , the redshift parameter @xmath59 is assumed to be known from the supernova candidate s host galaxy , @xmath93 , with an associated uncertainty of @xmath94 .
we consider the range of redshifts from 0 to 1.7 , and assume two representative possibilities for @xmath94 , 0.005 ( which might be obtained through a spectroscopic analysis of the host galaxy s spectrum ) and 0.1 ( obtained through a photometric analysis ) .
therefore , @xmath95 putting everything together , we obtain the full bayes factor : @xmath96 where @xmath97 represents the sums and integrations over the parameters in @xmath69 ( depending on whether they are discrete or continuous ) .
now the calculation of eqn .
[ eqn : exact_r ] requires performing @xmath98 integrations over @xmath99 in the denominator . for a large number of filters ( say ,
@xmath100 8) and many measurements , it is nearly impossible to do this calculation in a reasonable amount of time with the required precision for @xmath99 without the use of techniques such as markov chain monte carlo integration methods . however , the number of integrations can be reduced to @xmath48 ( the number of filters ) if we allow for measurement - to - measurement variations in @xmath19 .
that is , if we allow @xmath101 @xmath102 can be brought inside the product over measurements .
this assumption is the equivalent of removing the knowledge that the colors between the measurements are known ( or in the poisson case , this is equivalent to having a separate multinomial for each measurement ) .
this effectively releases the constraint on colors between measurements .
technically , the effect of this assumption should be to sweep more candidates which look less like sne ia into the sn ia hypothesis , so if one s sample contains `` anomalous '' candidates that very closely mimic sne ia one might reconsider this conjecture . with eqn .
[ eqn : b ] , eqn.[eqn : r_final ] becomes @xmath103
in order to check the performance of the method proposed above , we simulate a dataset closely mimicking one that could be obtained by a possible jdem space - based mission .
the mission is based on a 2-m class telescope , and is capable of taking multi - band photometric data in the wavelength range from 0.3 to 1.7 @xmath104 m .
photometric data are assumed to be taken every 4 days in the observer frame , with an exposure time of 1200 seconds .
note that this study is not meant to test the performance of any particular jdem mission ; we simply test the performance of the method assuming a fairly generic , plausible jdem .
we create a number of supernova candidates of a given type using spectral templates from @xcite for type ia s , and from p. e. nugent for non-ias.nugent/nugent_templates.html . ]
the date of explosion for a given candidate is chosen randomly within the confines of a 3 year mission timeline , and supernova candidate properties are generated according to the probability distribution functions in @xmath86 ( see eqn .
[ eqn : thetas ] ) .
the supernova light curves are then realized using a simple aperture exposure time calculator .
we generate supernova candidates of types ia , ibc , ii - p , and iin .
we assume that the intrinsic rest - frame @xmath105-band magnitudes follow a gaussian distribution , with the mean and standard deviations obtained from @xcite .
in particular , the mean and standard deviation are taken to be @xmath106 mags . for type ia s ; @xmath107 mags . for type ibc s ; @xmath108 mags . for type iin s and @xmath109 mags . for type ii - p s .
we also generate `` anomalous '' objects by creating fake unimodal light curves ( light curves that rise and fall with a single maximum , but are otherwise random ) .
these light curves are assigned 1% errors in each broadband filter considered . to get a better feel for the simulated dataset , fig .
[ fig : sn ] shows the signal - to - noise ratios at maximum light as a function of redshift for the simulated sne ia in the filter that most closely matches the rest - frame @xmath105-band .
= 3.0 in once the light curves are simulated , we select a subset of them in a limited number of broadband filters .
the chosen filters must include those that most closely match the rest - frame @xmath105- and @xmath110- bands for a given candidate .
the reason why we place a particular emphasis on these bands is because they correspond to a wavelength range where sne ia are particularly well modeled . to limit the computational time ,
we only consider 50 consecutive photometric measurements ( the actual available number of measurements varies evenly from 0 to over 350 ) .
we also require that there be at least one measurement with a signal - to - noise ratio @xmath111 5 in the filter most closely corresponding to the rest - frame @xmath105-band for a given candidate .
any space - based dark energy mission that extends to at least a year and uses sne ia as a dark energy probe will satisfy these requirements ( in fact , every jdem mission currently on the market does ) .
a number of tests are used to check the performance of the method .
first , we calculate the bayes factor , @xmath8 ( eqn .
[ eqn : r_final2 ] ) , for a sample of simulated sne ia and a sample containing unimodal `` fake '' light curves .
the unimodal light curves for a given `` object '' peak at the same time in all the filter bands . to give a sense of their color distribution , fig .
[ fig : fake_colors ] shows the fake objects colors for the 3 lowest wavelength filters bands ( the first filter covers the range of 0.32 - 0.47 @xmath104 m ; the second , 0.41 - 0.56 @xmath104 m ; and the third , 0.49 - 0.68 @xmath104 m ) .
= 3.0 in this test allows us to test the discrimination between sne ia and objects that are not supernovae of any kind .
figure [ fig : fakes ] shows that @xmath112 is predominantly positive for sne ia and negative for the random , unimodal data , meaning that @xmath113 .
this means that the method behaves as expected , discriminating between random data and true supernovae 100% of the time .
= 3.0 in we also compute the bayes factor for a sample of simulated type ibc s , type ii - p s , and type iin s .
note that we do not have to include any information about the expected light curves for type ibc s , type iin s , or type ii - p s to compute @xmath8 .
figure [ fig : iaandii - p ] shows the comparison of @xmath112 for the case of measurements in 2 filters ( left column ) and 4 filters ( right column ) .
we consider the case of a 0.1 error on the redshifts ( top row ) and a 0.005 error on the redshift ( bottom row ) . for this comparison , we assume that the errors on the supernova fluxes are realistic ( that is what a jdem mission described above would be expected to obtain ) .
@xmath114 several interesting conclusions can be drawn from fig .
[ fig : iaandii - p ] .
first , it is apparent that the method does discriminate between sne ia and the other types , although log@xmath8 tend to be larger than 0 ( so that @xmath115 ) because sne ia are far more similar to other supernovae than they are to anything else .
it should be noted that the values of @xmath8 tends to be quite large .
this is due to our use of a large number of measurements ( @xmath116 50 ) in a number of filters , which ensures that a candidate is either very much identified as an sn ia - like candidate or not .
second , as expected , the discrimination between sn ia s and non - ia s increases with more information ( 4 filters vs. 2 filters ) and/or with better prior knowledge ( _ i.e. _ , smaller errors on the measured redshift ) .
third , the discrimination is somewhat worse at high redshifts ( @xmath111 @xmath1161 ) , as we move into a domain of less precise data and less certain models ; but it is still good enough for @xmath8 to be used as a first - pass sn ia classifier .
additionally , at least some plausible jdems `` sculpt ''
their expected sne ia distribution so that it peaks at @xmath59 @xmath116 0.7 @xcite .
one might ask how @xmath8 would be affected if the templates used had incorrect colors for the sn ia hypothesis . in order to answer this question , we generated sn ia candidates with @xmath117 but use only the no - extinction templates when calculating @xmath8 .
figure[fig : wrongdust ] shows the distributions of @xmath8 for the cases where the extinction parameters in the data and templates are matched and mis - matched .
as expected , @xmath8 decreases when the extinction in the templates does not match that in the data ; that is , the sne ia look more similar to non - sne ia .
= 3.0 in we further check the behavior of @xmath8 by varying the errors on the fluxes of the simulated sne ia to ensure that @xmath8 changes in the right direction .
this check , always a good idea for a newly introduced statistic , is particularly important for this bayes factor , which makes use of improper priors that can lead to non - intuitive behavior ( @xcite ) .
figure [ fig : errs ] shows the distribution of @xmath112 for the `` nominal '' flux errors and for flux errors artificially increased and decreased by a factor of 2 . increasing the flux errors shifts the distribution to the left ( _ i.e. _
, the discrimination power decreases ) , while decreasing the flux errors shifts it to the right ( _ i.e. _ , the discrimination power increases ) .
this is the expected behavior for a correctly computed @xmath8 .
= 3.0 in so far , we have assumed no prior knowledge of sne models that may contribute to the set of observed non - sne ia .
we showed that the bayes factor described above is capable of discriminating between sne ia and non - sne ia as well as random unimodal light curves that mimic anomalous candidates .
this discrimination , which does not require either the knowledge of a complete set of objects that can mimic an sn ia signal or the knowledge of the possible behavior of anomalous non - supernova objects that can contaminate an sn ia signal , is good . as fig.[fig : iaandii - p ] shows , we could simply define a polynomial cut on @xmath8 as a function of @xmath8 and have a very good discriminant .
however , one might be interested in considering a bayes factor for which all _ known _ non - sn ia candidates would have @xmath113 .
first , it is simply better to have a @xmath8 that behaves intuitively .
second , it is obvious that including more prior knowledge ( _ i.e. _ , the knowledge of what can potentially mimic an sn ia signal ) , can only sharpen the discrimination between sne ia and non - sne ia .
third , and more importantly , it is a good idea to have a discriminant such that @xmath8 @xmath118 1 when the candidate is more likely to be an sn ia than not , and @xmath8 @xmath111 1 otherwise .
this allows one to use a formalism similar to that described in @xcite in order to set thresholds on @xmath8 with meaningful , pre - determined type i and ii error rates . explicitly including the knowledge about the behavior of non - sne ia into eqn .
[ eqn : general ] , we define : @xmath119 where the denominator now accounts for the probability that the observed photometry can come from a type ii - p supernova , @xmath120 ; a type ibc supernova , @xmath121 ; or from a type iin supernova , @xmath122 .
@xmath123 is equivalent to the denominator in eqn.[eqn : r_final2 ] .
note that the prior terms in eqn .
[ eqn : general2 ] are such that @xmath124 in other words , there is an equal probability of measuring a type ia supernova , a type ibc supernova , a type ii - p supernova or some other object denoted as `` anything '' .
it is of course trivial to introduce relative rates if they are known ; however , it is immaterial for our purpose , which is demonstrating the performance of the method .
the non - type ia probabilities are calculated in exactly the same way as those for type ia s , using the available models for the corresponding types .
the distribution of @xmath125 vs. redshift in shown in fig .
[ fig : iaandii - ppriorknowledge ] , for the case of 2 filters ( left column ) and 4 ( right column ) filters with 0.005 flux errors ( bottom row ) , and 0.1 errors ( top row ) on the redshift .
randomly generated , uni - modal data all have large negative values @xmath126 s that dwarf the scales on these figures .
the errors on the flux are those expected for a jdem mission described above . compared fig .
[ fig : iaandii - ppriorknowledge ] to fig.[fig : iaandii - p ] , it is clear that not only has the discrimination between sne ia and everything else increased , but also the sne ia generally have @xmath127 and other candidates have @xmath128 .
this is the desired behavior for @xmath129 .
@xmath130 a number of features of fig .
[ fig : iaandii - ppriorknowledge ] are similar to those apparent in fig .
[ fig : iaandii - p ] , such as the increase in the discrimination power when more and/or better information becomes available . as we explained in section [ section : intro ] , existing bayesian - based methods of supernova classification assume a finite set of possible objects that can mimic an sn ia signal . to demonstrate
the danger of this limiting assumption , we use our unimodal fake light curves and calculate bayes factors defined as : @xmath131 and @xmath132 figure[fig : naive ] shows that the fake objects can be mis - identified as supernovae of types other than ia ( in particular , as type ii - p s ) , as well as sne ia : there are candidates with @xmath133 @xmath111 1 .
this further demonstrates the need for a general a formalism that is capable of discriminating a certain type of supernovae ( most often type ia s ) from _ anything else_. = 4.0 in + = 4.0 in
we have introduced a new photometric supernova classification scheme that uses a bayes factor based on color .
the proposed method is fundamentally different from previous supernova classification methods including our own @xcite because it allows one to discriminate not only between supernovae of different types but also between supernovae and `` anomalous '' objects .
it has a number of definite advantages over many existing techniques for selecting sne ia out of a pool of supernova candidates .
the main one is that it does not pre - suppose any prior knowledge about the objects that could potentially mimic a type ia signal .
it can thus be used as a very good first - pass type ia classifier . with the current poor knowledge of the behavior of non - type ia supernovae , especially at high redshifts , and the expected dramatic increase in the discoveries of new , as yet unknown classes of transient astronomical objects
, this feature of the method will be invaluable for future supernova surveys .
this is not , however , an excuse not to obtain as much information about non - type ia supernovae as one possibly can , as evidenced by the advantage of computing @xmath134 , which includes information about the light curves of type ia , ibc , iin , and ii - p supernovae ( obviously , more information means a better performance ) .
another principal advantage of the proposed method is that if the bayes factor described in section [ section : results ] is used as a discriminant , the only supernova models that are required are those for sne ia , which are the best studied and most complete of all the supernova types .
one may argue that the same is true for a @xmath0-based method .
however , @xmath0 methods suffer from a number of problems described in section[section : intro ] , the least of which is that for the case of data with large uncertainties a given supernova candidate will appear to agree with many possible supernova type hypotheses , with one necessarily giving the `` best '' @xmath0 ( however insignificant the difference between this best @xmath0 and the @xmath0 s from the other fits may be ) .
the bayes factor , on the other hand , would be of order 1 , reflecting ambiguity between the hypothesis that the candidate is an sn ia and that it is anything else
. finally , as it is a bayesian approach , it accounts for all systematic and statistical uncertainties of the measurements in all their forms .
we demonstrated the method on a simulated dataset expected from a possible future large - scale space - based jdem .
it must be noted that most existing data sets include primarily sne ia ; at the present time large datasets with exotic supernovae that would allow us to perform a statistically meaningful study of the method s performance are not available ( although we are planning on exploring the application of this method to an existing dataset of non - standard supernovae in a future publication ) .
nevertheless , a number of interesting conclusions can be drawn from the simulation studies described above : * the method provides a 100% discrimination between sne ia and unimodal random data .
this is encouraging , since many transient objects ( such as active galactic nuclei ) that are sometimes mistaken for supernovae tend to have a rather erratic behavior , deviating far more from an sn ia light curve than the simple random unimodal data .
more fundamentally , the discrimination also shows that @xmath8 and @xmath134 behave as expected . *
the discrimination between type ia s and other supernova types is near 100% at low redshifts when calculating log@xmath8 . at higher redshifts , fig .
[ fig : iaandii - p ] shows that that a `` straight line '' cut on @xmath112 would render a reasonably high purity and efficiency for sne ia .
alternatively , one could invent a more sophisticated cut ( _ e.g. _ , a polynomial ) .
the discrimination is practically 100% at all redshifts if log@xmath134 is used ( _ i.e. _ , when information about the behavior of type ii - p , type ibc , and type iin supernovae is included in calculating @xmath3 ) . *
the method performs better when more information about the data becomes available . for example , the discrimination between sne ia and non - sne ia increases dramatically when the number of filters goes from 2 to 4 .
this shows that , despite the use of improper priors , the bayes factor behaves properly . *
increasing the precision on the redshift of the supernova candidates also improves the discrimination between type ia s and non - sne ia . *
it is important that the data and the models used have a good match in terms of expected colors .
for example , if the extinction assumptions in the data are different from those in the data , real sne ia are more likely to be classified as anomalous objects .
* figures[fig : iaandii - p ] and [ fig : iaandii - ppriorknowledge ] indicate that there is sufficient separation between various supernova candidates , so that this method could be used for classification in the strict sense of giving a type ( _ e.g. _ , a type ia , a type ibc , _ etc . _ ) to each candidate , along with some associated probability .
the method does not make use of magnitudes .
one might see this as an advantage if one does not entirely trust the distribution of intrinsic magnitudes .
insertion of absolute magnitudes is possible , but the formulation of @xmath8 ( @xmath134 ) becomes inelegant and requires knowledge of an upper limit of intrinsic magnitudes of anomalous candidates that , by definition , have not been observed .
it is indeed fortunate that color alone is sufficient to classify supernovae , allowing for a simple and elegant solution for @xmath8 ( @xmath134 ) .
there is another reason not to include magnitudes into the formalism , at least at the present time .
the distributions of the intrinsic magnitudes for non - sne ia are at the moment rather poorly known @xcite .
in fact , very little is known about high - redshift non - sne ia ( for example , @xcite remains the only measurement of the non - sne ia rates to redshifts of @xmath116 1 ) .
there exists a very real need to measure the properties of non - sne ia supernovae with more precision , a task that is ideally suited for existing and planned large - scale supernova surveys .
it is also important to note the computational challenges in calculating @xmath8 ( @xmath134 ) .
the number of filters used in the calculation depends on the precision needed for the integration of @xmath135 .
it was found that for 4 filters about 150 integration points for @xmath136 were needed for a precise calculation of the denominator , @xmath3 .
this was found by increasing the number of integration points until the value of the denominator became stable . in order to complete the computations in a reasonable amount of time
, it was necessary to perform the calculations of the bayes factor for many candidates in parallel .
the computational feasibility also depended on approximating @xmath137 , as was discussed in section [ section : details ] , so as to reduce the number of integrations from @xmath138 ( where @xmath47 is the number of flux measurements and @xmath48 is the number of filters ) to @xmath48 .
this effectively required that we gave up information about the supernova colors measured in the various filters and allowed the colors to vary measurement - to - measurement .
one might be concerned that this approximation would in fact allow for a greater diversity in what is considered an sn ia that is , in general objects would have a greater chance of faking an sn ia . however , in our studies we found that , at least with the level of precision of the models and simulated data used , the calculated bayes factor provided desired discrimination between sne ia and other objects , leading us to believe that this is not a significant effect .
we also point out that our proposed technique is general enough to be used for objects other than supernovae .
for example , @xcite propose a bayesian classifier to differentiate quasars and stars , defined as : @xmath139 where @xmath140 is a candidate s position in a 4-dimensional color space .
although the likelihoods , @xmath141 and @xmath142 are obtained using a training sample derived from real data , new data inevitably bring new objects which could be accounted for by inserting an `` anomaly '' term similar to @xmath143 in eqn.[eqn : general2 ] .
this term would sweep up those objects that do not conform to the existing models for quasars or stars in the same way that @xmath143 accounts for anomalous supernova candidates ; its exact form would depend on the nature of the statistical fluctuations in the data .
note that while our method relies only on photometric information about supernovae , some proposed future space - based dark energy missions do plan on obtaining the spectrum of every candidate .
one possible scenario would be to obtain the spectrum of a candidate provided that a ) it is highly likely to be an sn ia , and b ) it is at its peak brightness . in this case
it is necessary to have reliable means to be able to tell whether or not a given candidate is most likely an sn ia or not based on its pre - maximum photometric measurements alone . in other words , it is important to have a _ trigger _ mechanism that would photometrically select candidates for possible spectroscopic follow - up .
our proposed bayes factor can be simply modified to allow for such a usage . assuming that one wants to trigger on supernovae before maximum brightness , the bayes factor becomes : @xmath144 where @xmath145 would only include supernova models with points before maximum light , while @xmath146 would only include those with points after maximum light .
after calculating this bayes factor , a cut would be made at , say , @xmath147 to choose those candidates that are likely to be sne ia and that have not yet reached maximum .
better still , one could use a _ sequential analysis _
technique @xcite to minimize the data required to make this decision while simultaneously controlling identification errors .
this is done by setting thresholds on @xmath148 based on pre - selected type i and type ii error rates
. the demonstration of the performance of a sequential analysis - based approach will be the subject of a future publication .
nc gratefully acknowledges support from a cottrell college science award from the research corporation and from nsf rui award # 0806877 , as well as from hamilton college .
bc is partially supported by grant # de - fg02 - 95er40893 from the u.s .
department of energy .
we thank steve young at hamilton college for computer technical support .
barbary , k. , _ et al .
_ , 2009 , apj , 690 , 2 , 1358 baron , e. , nugent , p. , branch , d. , hauschildt , p. , 2004 ,
apj , 616 , 91 berger , j. , and pericchi , l. , 2001 , _ model selection _ , in institute of mathematical statistics lecture notes , ed .
p. lahiri , monograph series vol .
38 edwards , a. w. f. , 1992 , _ likelihood _ , johns hopkins university press hamuy , m. & pinto , p. a. 2002 , apj , 566 , l63 hsiao , e. y. , _ et al .
_ , 2007 , apj , 663 , 1187 gradshteyn , i. s. , & rhyzik , i. m. , 2000 , _ table of integrals , series and products _ , edited by a. jeffrey and d. zwillinger , academic press , new york , 6th edition guy , j. _ et al .
_ , 2007 , a&a , 466 , 11 jha , s. , riess , a. g. , & kirshner , r. p. , 2007 ,
apj , 659 , 122 jaynes , e. t. , and bretthorst , g. l. ( eds ) , 2003 , _ probability theory : the logic of science _ , cambridge university press jeffreys , h. , 1961 , _ theory of probability _ , third edition , oxford university press johnson , b. , and crotts , a. , 2006 , aj , 132 , 756 kuznetsova , n. , and connolly , b. m. , 2007 , apj , 659 , 530 nugent , p. et al .
2006 , apj , 645 , 841 perlmutter , s. , _ et al .
_ 1997 , apj , 483 , 565 poznanski , d. , _ et al .
_ , 2002 , pasp , 114 , 833 poznanski , d. , maoz , d. , and gal - yam , a. , 2007 , aj , 134 , 1285 poznanski , d. , _ et al .
_ , 2009 , apj , 694 , 2 , 1067 riess , a. g. , _ et al .
_ , 2004 , apj , 600 , l163 richards , g. t. , _ et al .
_ 2004 , astrophys.j.suppl . ,
155 , 257 richardson , d.,_et al .
_ 2002 , aj , 123 , 745 sellke , t. , bayarri , m. j. , and berger , j. o. , 2001 , _ the american statistician _ , 55 , 1 , 62 sullivan , m. , _ et al .
_ , 2006 , aj , 131 , 960 sullivan , m. , _ et al .
_ 2006 , apj , 648 , 868 wald , a. 1945 , ann .
stat . , 16 , 117 . 1947 , _ sequential analysis _ , john wiley and sons | upcoming large - scale ground- and space- based supernova surveys will face a challenge identifying supernova candidates largely without the use of spectroscopy . over the past several years
, a number of supernova identification schemes have been proposed that rely on photometric information only .
some of these schemes use color - color or color - magnitude diagrams ; others simply fit supernova data to models .
both of these approaches suffer a number of drawbacks partially addressed in the so - called bayesian - based supernova classification techniques . however , bayesian techniques are also problematic in that they typically require that the supernova candidate be one of a known set of supernova types .
this presents a number of problems , the most obvious of which is that there are bound to be objects that do not conform to any presently known model in large supernova candidate samples .
we propose a new photometric classification scheme that uses a bayes factor based on color in order to identify supernovae by type .
this method does not require knowledge of the complete set of possible astronomical objects that could mimic a supernova signal .
further , as a bayesian approach , it accounts for all systematic and statistical uncertainties of the measurements in a single step . to illustrate the use of the technique ,
we apply it to a simulated dataset for a possible future large - scale space - based joint dark energy mission and demonstrate how it could be used to identify type ia supernovae .
the method s utility in pre - selecting and ranking supernova candidates for possible spectroscopic follow - up _
i.e. _ , its usage as a supernova trigger will be briefly discussed . |
many mathematicians have already studied generalized topological spaces in the sense of csszr s article @xcite . although our article does not concern csszr s style generalization of topologies , we thank t. kubiak for turning our attention to the fact that generalizations of topologies such that it is not assumed that finite intersections of open sets are open , appeared in @xcite , earlier than in @xcite .
we are interested in the notion of a generalized topological space in the sense delfs and knebusch , introduced in @xcite .
unfortunately , only several articles about it have been published so far .
we use the abbreviation gts for a generalized topological space in the sense of delfs and knebusch .
contrary to simple generalizations of topological spaces given , for example , in @xcite and @xcite , the notion of a gts in the sense of delfs and knebusch is rather complicated since it originates from a categorical concept of grothendieck topology ( cf .
@xcite , @xcite ) .
the present work is the first study of problems relevant to quasi - metrizability in the class of delfs and knebusch gtses under the basic set - theoretic assumption of * zf * and its consistency . to avoid misunderstandings ,
let us make it more precise what * zf * is in this article .
since part of our results concern proper classes in category theory , while it is disturbingly assumed in @xcite that proper classes do not exist ( cf . pages 14 and 34 of @xcite ) , so far as * zf * is concerned , we follow @xcite and assume the existence of a universe @xmath0 ( cf . pages 22 and 23 of @xcite ) .
sets in the sense of @xcite are called _ totalities _ or _ collections_. a _ class _ is a collection @xmath1 .
elements @xmath2 are called @xmath0-small sets in @xcite . we denote by * zf * the system of axioms which consists of the existence of a universe @xmath0 and axioms 0 - 8 from pages 9 - 10 of @xcite for sets in the sense of @xcite
. this system does not contain the axiom of choice . from now on ,
a totality @xmath3 will be called a _ set _ if and only if @xmath3 is a @xmath0-small set .
a _ proper class _ is a class @xmath3 such that @xmath4 .
our notation concerning other set - theoretic axioms independent of * zf * that are used in this article is the same as in @xcite .
we clearly denote the results that are obtained not in @xmath5 but under @xmath6 where @xmath7 is the axiom of countable choice ( cf .
definition 2.5 of @xcite ) . of course
, not all axioms of * zf * are needed to deduce some results .
for instance , our results in * zf * that do not involve proper classes , can be deduced from the standard system of axioms 0 - 8 given on pages 9 - 10 of @xcite .
let us recall the following definition equivalent to the notion of a gts in the sense of delfs and knebusch : a _ generalized topological space _ in the sense of delfs and knebusch ( abbreviation : gts ) is a triple @xmath8 where @xmath9 is a set for which @xmath10 , while @xmath11 and the following conditions are satisfied : 1 .
if @xmath12 and @xmath13 is finite , then @xmath14 and @xmath15 ( where @xmath16 ) ; 2 . if @xmath17 and @xmath18 , then @xmath19 ; 3 .
if @xmath17 and , for each @xmath20 , we have @xmath21 such that @xmath22 , then @xmath23 ; 4 . if @xmath12 and @xmath24 are such that @xmath25 and , for each @xmath26 there exists @xmath27 such that @xmath28 , then @xmath17 ; 5 .
if @xmath17 , @xmath29 and , for each @xmath27 , we have @xmath30 , then @xmath31 .
if @xmath8 is a gts , then @xmath32 and , therefore , we can identify the gts with the ordered pair @xmath33 ( cf .
@xcite , @xcite ) .
if this is not misleading , we shall denote a gts @xmath33 by @xmath9 . in our approach to the problem of how to define a quasi - metrizable gts , we apply bornologies . according to @xcite , a _
bornology _ in a set @xmath9 is a non - empty ideal @xmath34 of subsets of @xmath9 such that each singleton of @xmath9 is a member of @xmath34 . a _ base of a bornology _
@xmath34 is a collection @xmath35 such that each member of @xmath34 is a subset of a member of @xmath36 .
a bornology is _ second - countable _ if it has a countable base .
bornological universe _ is an ordered pair @xmath37 where @xmath38 is a topological space and @xmath34 is a bornology in @xmath9 ( cf .
definition 1.2 of @xcite ) .
a _ generalized bornological universe _ is an ordered pair @xmath39 where @xmath40 is a gts and @xmath34 is a bornology in the set @xmath9 . a _ quasi - pseudometric _ on a set @xmath9 is a function @xmath41 such that , for all @xmath42 , @xmath43 and @xmath44 .
a quasi - pseudometric @xmath45 on @xmath9 is called a _ quasi - metric _ if , for all @xmath46 , the condition @xmath47 implies @xmath48 ( cf .
@xcite , @xcite ) .
let @xmath45 be a quasi - pseudometric on @xmath9 .
the _ conjugate _ of @xmath45 is the quasi - pseudometric @xmath49 defined by @xmath50 for @xmath51 .
the @xmath45-_ball with centre @xmath52 and radius @xmath53 _ is the set @xmath54 .
for a set @xmath55 and a number @xmath56 , the _
@xmath57-neighbourhood of @xmath58 with respect to @xmath45 _ is the set @xmath59^{\delta}_d=\bigcup_{a\in a}b_d(a,\delta)$ ] .
the collection @xmath60 is the _ topology in @xmath9 induced by @xmath45_. the triple @xmath61 is the _ bitopological space associated with @xmath45_. [ cf .
definition 1.5 of @xcite ] let @xmath45 be a quasi-(pseudo)metric on a non - empty set @xmath9 and let @xmath58 be a subset of @xmath9 .
then : 1 .
@xmath58 is called
_ @xmath45-bounded _ if there exist @xmath52 and @xmath62 such that @xmath63 ; 2 .
if @xmath58 is not @xmath45-bounded , we say that @xmath58 is _ @xmath45-unbounded _ ; 3 .
@xmath64 is the collection of all @xmath45-bounded subsets of @xmath9 .
in addition , if @xmath65 , one can treat @xmath66 as the unique quasi - pseudometric on @xmath9 and , of course , the empty set should be also called @xmath45-bounded in the case of the empty space .
it was shown in example 1.6 of @xcite that , for a quasi - metric @xmath45 on @xmath9 , a set @xmath55 can be both @xmath45-bounded and @xmath49-unbounded .
definition 1.7 of @xcite ] a bornological universe @xmath67 is called _ quasi-(pseudo)metrizable _ if there exists a quasi-(pseudo)metric @xmath45 on @xmath9 such that @xmath68 and , moreover , @xmath34 is the collection of all @xmath45-bounded sets . for a collection @xmath69 of subsets of a set @xmath9 ,
we denote by @xmath70 the weakest among all topologies in @xmath9 that contain @xmath69 . for a gts @xmath8
, we call the topological space @xmath71 the _ topologization of the gts _ @xmath9 ( cf .
@xcite ) .
suppose that @xmath39 is a generalized bornological universe .
then we say that the gts @xmath9 is _
@xmath34-(quasi)-(pseudo)metrizable _ or _
( quasi)-(pseudo)metrizable with respect to @xmath34 _ if the bornological universe @xmath72 is ( quasi)-(pseudo)metrizable .
a _ bornological biuniverse _ is an ordered pair @xmath73 where @xmath74 is a bitopological space and @xmath34 is a bornology in @xmath9 . to a great extent , the present work is a continuation of @xcite .
therefore , let us use the terminology of @xcite .
definition 1.3 of @xcite ] let @xmath75 be a bitopological space .
a bornology @xmath34 in @xmath9 is called @xmath76__-proper _ _ if , for each @xmath77 , there exists @xmath78 such that @xmath79 .
if @xmath80 and the bornology @xmath34 is @xmath81-proper , we say that @xmath34 is @xmath82__-proper__. let us formulate our first ( quasi)-(pseudo)metrization theorem in the class of gtses which follows from the results of @xcite .
the notion of a @xmath83-characteristic function of a bornology is introduced in definition 4.4 of @xcite . for every gts @xmath84 and a bornology @xmath34 in @xmath9 , the following conditions are equivalent : 1 .
@xmath85 is a ( quasi)-(pseudo)metrizable generalized bornological universe ; 2 .
there exists a topology @xmath86 on @xmath9 such that @xmath87 is ( quasi)-(pseudo)metrizable and @xmath34 is a @xmath88-proper bornology with a countable base ; 3 .
there exists a topology @xmath86 on @xmath9 such that @xmath87 is ( quasi)-(pseudo)metrizable and @xmath34 admits a @xmath88-characteristic function ; 4 .
there exists a ( quasi)-(pseudo)metric @xmath45 on @xmath9 such that @xmath89 and , simultaneously , @xmath34 has a base @xmath90 with the following property : @xmath91^{\delta}_d\subseteq b_{n+1}.\ ] ] in section 2 , we discuss natural bornologies in every gts : the small sets , the relatively compact sets and the relatively admissibly compact sets .
other ( quasi)-(pseudo)metrization theorems for gtses are given in section 3 .
the main theorem of section 3 ( theorem 3.5 ) gives a necessary and sufficient condition for a gts with its locally small partial topologization to be ( quasi)-(pseudo)metrizable with respect to the bornology of small sets .
moreover , a notion of a weakly locally small gts is introduced and a non - trivial example of a weakly locally small but not locally small gts is constructed in every model for @xmath6 in section 3 .
applications of the ( quasi)-(pseudo)metrization theorems to numerous examples are shown in section 4 . finally , in section 5 , we define several new categories relevant to this work and we prove that the newly defined categories are all topological constructs .
so far as gtses are concerned , we use the terminology of @xcite , @xcite , @xcite and @xcite . if @xmath92 and @xmath93 are gtses , then : 1 .
a set @xmath94 is called _ open _ in the gts @xmath9 if @xmath95 ; 2 .
the collection @xmath96 is the _
generalized topology _ in @xmath9 ; 3 .
an _ admissible open family _ in the gts @xmath9 is a member of @xmath96 ; 4 .
a mapping @xmath97 is @xmath98-_strictly continuous _
( in abbreviation : strictly continuous ) if , for each @xmath15 , we have @xmath99 .
we denote by @xmath100 the category of gtses as objects and strictly continuous mappings as morphisms .
the category @xmath100 is a subcategory of the category of grothendieck sites and their morphisms ( cf .
@xcite , p. 223 ) .
several other categories , relevant to @xmath100 and bornologies , are defined in section 5 . while we know that @xmath100 is a topological construct ( cf .
theorem 2.2.60 of @xcite and theorem 4.4 of @xcite ) , the following problem is open : is @xmath100 isomorphic to a category of type @xmath101-@xmath102 for some quantale @xmath103 and monad @xmath104 ( cf .
definitions iii.1.6.1 and iii.1.6.3 in @xcite ) ?
in this section , we consider natural bornologies in gtses .
if @xmath105 is a subset of a set @xmath9 , then we say that a family @xmath106 is _ essentially finite on _
@xmath105 if there exists a finite @xmath107 such that @xmath108
. if @xmath109 is a gts , then a set @xmath110 is called _ small in the gts @xmath9 _ if each family @xmath17 is essentially finite on @xmath105 .
the collection of all small sets of a gts @xmath9 is a bornology in @xmath9 ( cf .
fact 2.2.30 of @xcite ) . for a gts @xmath9 ,
the _ small bornology _ of @xmath9 is the collection @xmath111 of all small sets in @xmath9 .
@xmath111 was denoted by @xmath112 in @xcite but , since we use the notation of @xcite , we have replaced @xmath112 by @xmath111 .
the bornology of all finite subsets of @xmath9 ( the smallest bornology of @xmath9 ) is denoted by @xmath113 .
if @xmath9 is a gts , we call a set @xmath55 _ admissibly compact _ in @xmath9 if , for each @xmath17 such that @xmath114 , there exists a finite @xmath115 such that @xmath116 . for a gts @xmath9 ,
the _ admissibly compact bornology _ of @xmath9 is the collection @xmath117 of all subsets of admissibly compact sets of the gts @xmath9 .
let @xmath9 be a gts .
we say that a set @xmath58 is _ topologically compact _ in @xmath9 if @xmath58 is compact in @xmath118 ( cf .
definition 3.2 of @xcite ) .
the compact bornology @xmath119 ( cf .
@xcite and @xcite ) will be called the _ compact bornology of the gts _ @xmath9 and it
will be denoted by @xmath120 . for every gts @xmath9 , the inclusion @xmath121 holds .
the following example shows that it can happen that @xmath122 and neither @xmath123 nor @xmath124 .
[ comp - born ] for @xmath125 , let @xmath126 be the natural topology in @xmath9 inherited from the usual topology of @xmath127 and let @xmath96 be the collection of all families @xmath12 such that @xmath13 is essentially finite on @xmath128 .
then , for @xmath129\times\{1\}$ ] and @xmath130 , we have @xmath131 and @xmath132 , while @xmath133 . for a set @xmath9 and a collection @xmath134 , we denote by @xmath135 the smallest among generalized topologies in @xmath9 that contain @xmath136 .
if @xmath137 , let @xmath138 be the collection of all essentially countable subfamilies of @xmath69 .
we recall that @xmath139 is the collection of all essentially finite subfamilies of @xmath69 ( cf .
@xcite-@xcite and @xcite ) .
assume @xmath38 is a topological space . that @xmath140 is a generalized topology in @xmath9 is true in @xmath5 .
on the other hand , that @xmath141 is a generalized topology in @xmath9 is true in @xmath6 .
it is unprovable in @xmath5 that , for every topological space @xmath38 , the collection @xmath141 is a generalized topology in @xmath9 .
namely , let @xmath142 be a model for @xmath143 where @xmath144 states that countable products on non - empty finite sets are non - empty ( cf . definition 2.9(3 ) of @xcite ) . in view of proposition 3.5 of @xcite , there exists in @xmath142 an uncountable set @xmath9 such that @xmath9 is a countable union of finite sets .
let @xmath145 . if @xmath141 were a generalized topology in @xmath9 , the family of all singletons of @xmath9 would belong to @xmath141 which is impossible , since @xmath9 is uncountable .
let us observe that , for the gts @xmath9 from example [ comp - born ] , the admissibly compact bornology of @xmath9 is generated by @xmath146 . that not every gts may share this property
is shown by the following example : for @xmath147 , let @xmath126 be the topology induced by the usual linear order in @xmath148 and let @xmath149
. then @xmath150 . in
what follows , for sets @xmath151 with @xmath152 and for @xmath153 , we use the notation @xmath154 from @xcite for the collection of all families @xmath155 where @xmath156 .
we want to describe @xmath157 more precisely in the case when @xmath158 . to do this
, we need the concept of a full ring of sets in @xmath159 that was of frequent use in @xcite .
namely , a _ full ring in @xmath159 _ is a collection @xmath160 such that @xmath161 , while @xmath162 is closed under finite unions and under finite intersections . for @xmath163 ,
let @xmath164 $ ] be the intersection of all full rings in @xmath159 that contain @xmath69
. for a set @xmath9 , let @xmath153 .
suppose that @xmath152 and that each family from @xmath136 is essentially finite on @xmath159 .
then the following conditions are satisfied : 1 .
@xmath165)=\\\text{essfin}(\bigcup\langle\psi\cap_2 y\rangle_y)=\text{essfin}(\bigcup\langle \psi\rangle_x)\cap_2 y$ ] ; 2 .
each family from @xmath166 is essentially finite on @xmath159 . by applying proposition 2.2.37 of @xcite to the mapping @xmath167
, we obtain the inclusion @xmath168 which , together with @xmath169 , implies @xmath170 . to prove @xmath169 ,
let us put @xmath171 ) , \mathcal{g}_2=\text{essfin}(\bigcup\mathcal{g}_0)$ ] and @xmath172 . obviously , @xmath173 and @xmath174 are generalized topologies in @xmath159 . by proposition 2.2.53 of @xcite ,
the collection @xmath175 is also a generalized topology in @xmath159 . since @xmath176 and @xmath177\subseteq \bigcup\mathcal{g}_0 $ ] , we have @xmath178 .
it follows from the inclusion @xmath179 that @xmath180 .
since @xmath181 is a full ring of subsets of @xmath159 , we get @xmath182 .
this completes our proof to @xmath169 .
if @xmath183 is a gts , then : 1 .
the _ partial topologization _ of @xmath183 is the gts @xmath184 where @xmath185 and @xmath186 ( cf .
definition 4.1 of @xcite ) ; 2 .
the gts @xmath9 is called _ partially topological _ if @xmath187 ( cf .
definition 2.2.4 of @xcite ) ; 3 .
@xmath188 is the category of all partially topological spaces and strictly continuous mappings , while the mapping @xmath189 is the _ functor of partial topologization _
defined by : @xmath190 for every gts @xmath9 and @xmath191 for every morphism in @xmath100 ( cf .
@xcite , @xcite , @xcite and definition 4.2 of @xcite ) .
[ bornologie - pt ] let @xmath9 be a gts .
then @xmath192 , @xmath193 and @xmath194 .
the equality @xmath193 and both the inclusions @xmath195 and @xmath194 are trivial .
let @xmath196 and let @xmath197 .
suppose that @xmath198 . since each family from @xmath136 is essentially finite on @xmath159 , we infer from proposition 2.12 that @xmath199 .
let @xmath200 be a full ring of subsets of a set @xmath9 .
then : 1 . for a collection @xmath201
, we define @xmath202 2 . for a topology @xmath82 in @xmath9 and for a bornology @xmath34 in @xmath9 ,
the _ gts induced by the bornological universe _
@xmath37 is the triple @xmath203 . in the light of the proof to proposition 2.1.31 in @xcite
, we have the following fact : suppose that @xmath37 is a bornological universe such that @xmath204 is a base for @xmath34
. then @xmath205 . for a ( quasi)-(pseudo)metric @xmath45 on a set @xmath9 , the triple @xmath206
will be called the _ gts induced by the ( quasi)-(pseudo)metric _ @xmath45 .
if @xmath45 is a quasi-(pseudo)metric on a set @xmath9 , then @xmath207 is a generalized topology in @xmath9 and @xmath208
let @xmath9 be a gts and let @xmath209 be either @xmath210 or @xmath211 , or @xmath212 .
then we say that @xmath9 is @xmath209-(quasi)-(pseudo)metrizable if @xmath9 is ( quasi)-(pseudo)metrizable with respect to @xmath213 . with proposition [ bornologie - pt ] in hand
, we can immediately deduce that the following proposition holds : let @xmath209 be either @xmath210 or @xmath212
. then the following are equivalent for a gts @xmath9 : 1 .
@xmath9 is @xmath209-(quasi)-(pseudo)metrizable , 2 .
@xmath214 is @xmath209-(quasi)-(pseudo)metrizable .
if @xmath9 is a gts , then the @xmath211-(quasi)-(pseudo)metrizability of @xmath214 is the ( quasi)-(pseudo)metrizability of @xmath214 with respect to @xmath215 , while the @xmath211-(quasi)-(pseudo)metrizability of @xmath9 is equivalent to the ( quasi)-(pseudo)metrizability of @xmath214 with respect to @xmath117 .
we do not know whether the @xmath211-(quasi)-(pseudo)metrizability of @xmath9 is equivalent to the @xmath211-(quasi)-(pseudo)metrizability of @xmath214 .
a gts @xmath196 is called : 1 . _
locally small _ if there exists @xmath17 such that @xmath216 and @xmath217 ( cf .
definition 2.1.1 of @xcite ) ; 2 .
_ weakly locally small _ if there exists a collection @xmath218 such that @xmath217 .
our next theorem says about the form of the partial topologization of an @xmath212-(quasi)-(pseudo)metrizable gts @xmath9 when @xmath214 is locally small .
suppose that @xmath183 is a gts such that its partial topologization @xmath219 is locally small .
then the following conditions are equivalent : 1 .
@xmath9 is @xmath212-(quasi)-(pseudo)metrizable ; 2 .
@xmath214 is induced by some ( quasi)-(pseudo)metric @xmath45 . in view of proposition [ bornologie - pt ]
, we have @xmath192 . in consequence , it it is obvious that if @xmath214 is induced by a ( quasi)-(pseudo)metric @xmath45 , then @xmath9 is @xmath212-(quasi)-(pseudo)metrizable .
assume that @xmath9 is @xmath212-(quasi)-(pseudo)metrizable and that @xmath45 is a ( quasi)-(pseudo)metric on @xmath9 such that @xmath220 and @xmath221 is the collection of all @xmath45-bounded sets .
since @xmath214 is locally small , it follows from proposition 2.1.18 of @xcite that @xmath214 is induced by @xmath45 .
if a gts @xmath9 is induced by a ( quasi)-(pseudo)metric , then @xmath9 is locally small and partially topological . 1 .
if @xmath9 is a locally small gts , then @xmath214 is locally small .
2 . if a gts @xmath9 is such that @xmath214 is locally small , then @xmath9 is weakly locally small .
a gts @xmath9 is weakly locally small if and only if @xmath214 is weakly locally small . in every model for @xmath6
, we are going to present a construction of an example of a weakly locally small gts @xmath9 such that @xmath214 is not locally small . for @xmath222 , we put @xmath223 and , for @xmath224 , assuming that the collection @xmath225 has been defined , we put @xmath226 where @xmath227 is the operator described in the proof of proposition 2.2.37 in @xcite .
then @xmath228 .
the symbols @xmath229 have the same meaning as in @xcite .
we shall use the terminology from definition 2.2.2 of @xcite . @xmath230 $ ] .
suppose that @xmath159 is an uncountable set .
for @xmath224 , we put @xmath231 .
let @xmath232 , @xmath233 and @xmath234 . the gts @xmath196 is weakly locally small and not small . if @xmath9 were locally small , then @xmath235 would be a subset of a small open set ( fact 2.1.21 in @xcite ) , so @xmath235 would be finite . hence , @xmath9 is not locally small .
we have @xmath236 and all the sets @xmath237 are small and open in @xmath238 @xmath239 , so the gts @xmath240 is locally small .
we put @xmath241 .
then @xmath242 is the generalized topology of @xmath214 . by proposition 2.12 , @xmath243 .
surprisingly , if @xmath7 holds , then @xmath214 is not locally small and , in consequence , @xmath244 . to prove this ,
let us assume @xmath6 .
it is easy to observe the following facts : + * fact 1 .
* @xmath245^{\leq\omega}\cup_1 \mathbf{sm}(x)$ ] .
+ * fact 2 .
* each @xmath246 is closed with respect to restriction : @xmath247 for @xmath55 .
( notice that @xmath248 and @xmath249 is the hardest case . )
+ for @xmath250 , let us consider the following property : @xmath251 : @xmath252 has an uncountable member and @xmath253^{\leq\omega}\cup_1 \mathbf{sm}(x)$ ] .
+ for @xmath224 , let @xmath254 be the statement : @xmath254 : if @xmath255 has @xmath251 , then @xmath252 is essentially finite on @xmath256 for some countable @xmath55 .
+ we are going to prove by induction that the following fact holds : + * fact 3 .
* @xmath254 is true for each @xmath257 .
let @xmath258 have property @xmath251 .
then , by fact 1 , @xmath259 and @xmath260 .
hence @xmath261 and @xmath262 holds .
suppose that @xmath254 is true .
the _ finiteness _ , _ stability _ and _ regularity _ induction steps from the proof of proposition 2.2.37 in @xcite are obvious ( cf .
definition 2.2.2 of @xcite ) . _ transitivity step .
_ let @xmath263 have property @xmath251 .
suppose that @xmath264 and @xmath265 are such that @xmath266 and , for each @xmath27 , we have @xmath267 . consider any @xmath27 .
if every member of @xmath268 is countable , then @xmath269^{\leq\omega}$ ] because @xmath7 holds and @xmath268 is essentially countable .
suppose @xmath268 has an uncountable member .
since @xmath268 has property @xmath270 , it follows from the inductive assumption that there is a countable set @xmath271 such that @xmath268 is essentially finite on @xmath272
. then @xmath273^{\leq\omega}\cup_1 \mathbf{sm}(x)$ ] and @xmath274 is uncountable .
the above implies that @xmath13 has property @xmath275 . by the assumption
, there is a countable @xmath55 such that @xmath13 is essentially finite on @xmath256 .
let @xmath276 be a finite family such that @xmath277 . for each @xmath278 ,
the set @xmath274 is countable or @xmath268 is essentially finite on @xmath279 .
this implies that there is a countable @xmath280 such that @xmath252 is essentially finite on @xmath281 .
_ saturation step .
_ suppose that there exists @xmath282 such that @xmath283 and , for each @xmath26 , there exists a non - empty @xmath284 . since @xmath253^{\leq\omega}\cup_1\mathbf{sm}(x)$ ] , we have @xmath285^{\leq\omega}\cup_1 \mathbf{sm}(x)$ ] .
since @xmath252 has an uncountable member and @xmath286 is essentially countable , also @xmath286 has an uncountable member and has property @xmath287 . by the inductive assumption
, there exists a countable @xmath288 such that @xmath286 is essentially finite on @xmath289 .
then @xmath252 is essentially finite on @xmath289 , too .
suppose that @xmath214 is locally small .
there exists @xmath290 such that @xmath291 and @xmath292 .
since @xmath9 is uncountable and @xmath252 is essentially countable , at least one member of @xmath252 is uncountable , so @xmath251 holds true . by fact 3 , there exists a countable @xmath280 such that @xmath252 is essentially finite on @xmath281 . then @xmath293 .
this is impossible by fact 1 .
from fact 3.7 , taken together with example 3.8 , we deduce the following corollary : in every model for @xmath6 , there exists a gts @xmath9 such that @xmath294 , while both @xmath9 and @xmath214 are simultaneously weakly locally small and not locally small .
we do not have a satisfactory solution to the following open problem : is it true in @xmath5 that if the partial topologization of a gts @xmath9 is locally small , then so is @xmath9 ?
suppose that @xmath196 is a gts and @xmath34 is a bornology in @xmath9 .
then the following conditions are equivalent : 1 .
the gts @xmath9 is ( quasi)-(pseudo)metrizable with respect to @xmath34 ; 2 .
the gts @xmath295 is @xmath212-(quasi)-(pseudo)metrizable and the collection @xmath296 is a base for @xmath34 .
assume that @xmath169 holds .
then , by theorem 4.7 of @xcite , the collection @xmath296 is a base for @xmath34 .
it follows from fact 2.16 that @xmath297 . in consequence , @xmath169 implies @xmath170 . on the other hand
, we can use fact 2.16 with both definitions 1.6 and 2.15 to infer that @xmath169 follows from @xmath170 .
suppose that @xmath39 is a generalized bornological universe where @xmath196 .
let us say that @xmath9 is _
strongly @xmath34-(quasi)-(pseudo)metrizable _ if there exists a ( quasi)-(pseudo)metric @xmath45 on @xmath9 such that @xmath34 is the collection of all @xmath45-bounded sets and @xmath298 $ ] . in connection with strong @xmath212-(quasi)-(pseudo)metrizability ,
let us pose the following open problem : find useful simultaneously necessary and sufficient conditions for a gts to be strongly @xmath212-(quasi)-(pseudo)metrizable .
a _ ( quasi)-(pseudo)metric gts _ is an ordered pair @xmath299 where @xmath300 is a
gts and @xmath45 is a ( quasi)-(pseudo)metric on @xmath9 such that @xmath301 . for a ( quasi)-(pseudo)metric gts @xmath299 and a bornology @xmath34 in @xmath9
, we say that @xmath299 is _ uniformly ( quasi)-(pseudo)metrizable with respect to @xmath34 _ or , equivalently , that @xmath34 is _ uniformly ( quasi)-(pseudo)metrizable with respect to @xmath45 _ if there exists a ( quasi)-(pseudo)metric @xmath302 on @xmath9 such that @xmath45 and @xmath302 are uniformly equivalent , and @xmath303 ( cf .
definition 6.3 of @xcite and definition 2.3 of @xcite ) .
the following proposition follows from theorem 6.5 of @xcite : suppose that @xmath299 is a ( quasi)-(pseudo)metric gts .
then a bornology @xmath34 in @xmath9 is uniformly ( quasi)-(pseudo)metrizable with respect to @xmath45 if and only if @xmath34 has a base @xmath90 such that , for some @xmath56 and for each @xmath224 , the inclusion @xmath304^{\delta}_d\subseteq b_{n+1}$ ] holds .
other conditions equivalent to uniform @xmath34-(quasi)-(pseudo)metrizability of ( quasi)-(pseudo)metric gtses can be deduced from the results of @xcite and from section 6 of @xcite .
for @xmath305 , let @xmath306 .
we denote by @xmath307 the natural topology of @xmath127 induced by the metric @xmath308 .
let @xmath309 stand for the compact bornology of @xmath310 .
the topology @xmath311 is called the _ upper topology _ on @xmath127 , while @xmath312 is called the _ lower topology _ on @xmath127 ( cf .
@xcite , @xcite ) .
the following collections : @xmath313 are simple examples of bornologies in @xmath127 . obviously , @xmath314 the topological space @xmath315 is not quasi - metrizable ( since it is not @xmath316 ) but it is quasi - pseudometrizable by @xmath317 . 1 . for the gts @xmath318 , one has @xmath319 .
this is why @xmath320 is both @xmath211- and @xmath212-quasi - pseudometrizable by @xmath321 .
2 . for the gts @xmath322
, we have @xmath323 .
hence @xmath324 is both @xmath211- and @xmath212-quasi - pseudometrizable by @xmath325 .
3 . the gts @xmath326 is equal to @xmath320 .
the gts @xmath327 is equal to @xmath324 . 5 .
the gts @xmath328 is not @xmath329-quasi - pseudometrizable because @xmath330 for each @xmath331 . here
@xmath332 is the collection of all sets @xmath333 such that every non - empty subset of @xmath58 has its largest element .
similarly , @xmath334 is not @xmath212-quasi - pseudometrizable . since @xmath335 ,
the gts @xmath334 is @xmath211-quasi - pseudometrizable by @xmath321 .
each of @xmath336 is @xmath210-quasi - pseudometrizable by @xmath321 .
let us use the real lines described in definition 1.2 of @xcite as part of our illuminating examples for the notions of ( uniform ) @xmath34-(quasi)-metrizability in the category @xmath100 . for @xmath337
, we put @xmath338 and @xmath339 moreover , we define @xmath340 .
let us observe that the metrics @xmath308 and @xmath341 are equivalent but not uniformly equivalent .
( i ) : : let @xmath342 be any generalized topology in @xmath127 such that @xmath343 is a metric gts .
+ we have @xmath344 and @xmath345 .
let us observe that , for a fixed @xmath56 , there exists @xmath346 such that if @xmath347 $ ] for @xmath348 with @xmath349 , then @xmath350^{\delta}_{d^+_n}$ ] .
this , together with proposition 3.16 , implies that @xmath351 is not uniformly quasi - metrizable with respect to @xmath341 .
( ii ) : : for the usual topological real line @xmath352 ( cf . definition 1.2(i ) of @xcite ) , we have @xmath353 and @xmath354 for each @xmath355 , so the gts @xmath352 is not @xmath212-quasi - metrizable and it is @xmath211-metrizable by @xmath308 . the metric gtses @xmath356 and @xmath357 are @xmath211-uniformly metrizable .
it follows from ( i ) that the metric gtses @xmath358 is @xmath359 are not uniformly @xmath211-quasi - metrizable .
( iii ) : : for the real lines @xmath360 and @xmath361 ( cf .
definition 1.2(iv)-(v ) of @xcite ) , we have @xmath362 and @xmath363 . the metric gtses @xmath364 and @xmath365 are both uniformly @xmath212-metrizable ; however , none of the metric gtses @xmath366 and @xmath367 is uniformly @xmath212-metrizable ( see ( i ) ) .
( iv ) : : for the real lines @xmath368 and @xmath369 ( cf
. definition 1.2(vii)-(viii ) of @xcite ) , we have @xmath370 and @xmath371 .
now , it is obvious that both the metric gtses @xmath372 and @xmath373 are uniformly @xmath211-metrizable by the metric @xmath341
. the gtses @xmath368 and @xmath369 are @xmath212-metrizable .
the metric gtses @xmath374 and @xmath375 are uniformly @xmath212-metrizable and uniformly @xmath211-metrizable by @xmath376 .
( v ) : : let us consider the gtses @xmath377 and @xmath378 ( cf .
definition 1.2(ii),(iii ) , ( vi ) and ( x ) of @xcite ) .
we have @xmath379 and @xmath380 .
the real lines @xmath377 and @xmath378 are @xmath212-metrizable by the metric @xmath381 and they are @xmath210-metrizable by the metric @xmath382 .
( vi ) : : the gts @xmath383 ( cf .
definition 1.2(ii ) of @xcite ) is strongly @xmath212-metrizable by @xmath381 .
a famous quasi - metrizable but non - metrizable tychonoff space is the sorgenfrey line , denoted here by @xmath384 . the topology @xmath385 of @xmath384 is the the right half - open interval topology in @xmath127 .
the space @xmath384 is quasi - metrizable by the quasi - metric @xmath386 defined , for @xmath387 , as follows : @xmath388 by example 4.12 of @xcite , it is worthwhile to notice that the topology @xmath385 is also induced by the quasi - metric @xmath389 defined , for @xmath387 , as follows : @xmath390 now , we may consider the following gtses , obtained from the sorgenfrey line , that are modifications of the gtses defined in definition 1.2 of @xcite : [ proste sorgenfreya ] we give names to the following sorgenfrey real lines : 1 . the _ usual topological sorgenfrey real line _
@xmath391 where @xmath392 all families of members of @xmath385 ; 2 .
the _ o - minimal sorgenfrey real line _
@xmath393 where @xmath392 essentially finite families of finite unions of right half - open intervals ; 3 .
the _ smallified topological _ ( or _ small partially topological _ ) _ sorgenfrey real line _
@xmath394 where @xmath392 essentially finite families of members of @xmath385 ; 4 .
the _ localized o - minimal sorgenfrey real line _
@xmath395 where @xmath392 locally essentially finite families of locally finite unions of right half - open intervals ; 5 .
the _ localized smallified topological sorgenfrey real line _
@xmath396 where @xmath392 locally essentially finite families of members of @xmath385 ; 6 .
the _ smallified localized o - minimal sorgenfrey real line _
@xmath397 where @xmath392 essentially finite families of locally finite unions of right half - open intervals ; 7 .
the _ localized at @xmath398 _ ( @xmath399 , resp . )
o - minimal sorgenfrey real line _
@xmath400 ( @xmath401 , resp . ) where @xmath392 locally essentially finite families of locally finite unions of right half - open intervals which , on the negative ( positive , resp . )
half - line , are essentially finite and consist of only finite unions of right half - open intervals ; 8 .
the _ localized at @xmath398 _ ( @xmath399 , resp . )
_ smallified topological sorgenfrey real line _ @xmath402 ( @xmath403 , resp . )
where @xmath392 locally essentially finite families of members of @xmath385 which are essentially finite on the negative ( positive , resp . )
half - line ; 9 .
the _ smallified localized at @xmath398 _ ( @xmath399 , resp . ) _ o - minimal sorgenfrey real line _
@xmath404 ( @xmath405 , resp . )
where @xmath392 essentially finite families of locally finite unions of right half - open intervals which are only finite unions of right half - open intervals on the negative ( positive , resp . ) half - line ; 10 .
the _ rationalized o - minimal sorgenfrey real line _
@xmath406 where @xmath392 essentially finite families of finite unions of right half - open intervals with endpoints being rational numbers or infinities . in this case , the topology @xmath407 differs from @xmath408 .
( * gtses from the sorgenfrey line*. ) 1 .
the gtses @xmath409 are both @xmath211- and @xmath212-quasi - metrizable by the quasi - metric @xmath410 defined as follows : @xmath411 2 .
the gtses @xmath412 are both @xmath211- and @xmath212-quasi - metrizable by @xmath413 , while the gtses @xmath414 are both @xmath211- and @xmath212-quasi - metrizable by @xmath415 .
the gtses @xmath416 are @xmath211- and @xmath212-quasi - metrizable by @xmath417 .
4 . it follows from theorem 1.9 that the gts @xmath391 is neither @xmath212-quasi - metrizable because @xmath418 is not a base for @xmath419 , nor @xmath211-quasi - metrizable ( see below ) .
since each relatively compact set in the sorgenfrey topology is countable , none of the sorgenfrey lines defined above is @xmath210-metrizable .
( * quasi - metric gtses from the sorgenfrey line*. ) we use the same notation as in example 4.4 . 1 .
the quasi - metric gtses @xmath420 , @xmath421 are uniformly @xmath422-quasi - metrizable by @xmath410 .
the quasi - metric gtses @xmath423 , @xmath424 are uniformly @xmath422-quasi - metrizable by @xmath386 ,
while the quasi - metric gtses @xmath425 , @xmath426 are uniformly @xmath422-quasi - metrizable by @xmath389 , 3 .
the quasi - metric gtses @xmath427 ,
@xmath428 , @xmath429 , @xmath430 are uniformly @xmath422-quasi - metrizable by @xmath431 .
4 . none of the quasi - metric gtses @xmath432
, @xmath433 is uniformly @xmath422-quasi - metrizable , where @xmath434 let us put @xmath435\times\{0\}$ ] and @xmath436 $ ] . for @xmath437\cap\mathbb{q}$ ] , let @xmath438 .
we consider the collection @xmath34 of all sets @xmath55 that have the property : there exists a finite @xmath439 such that @xmath440 . 1 .
let @xmath441 be the euclidean metric on @xmath9 .
then , for each @xmath77 , we have @xmath442 , so , for every topology @xmath86 in @xmath9 , the bornology @xmath34 is not @xmath443-proper . in consequence , by theorem 1.9 , the gts @xmath444 is not @xmath212-quasi - metrizable .
we define another metric @xmath302 on @xmath9 as follows . for @xmath445 $ ] and @xmath446 with @xmath447
, we put @xmath448 and @xmath449 .
then , for each @xmath450 and for any @xmath451 $ ] with @xmath452 , we have @xmath453\in\mathcal{b}$ ] .
since there does not exist @xmath77 such that @xmath454 , we deduce from theorem 1.9 that the gts @xmath455 is not @xmath212-quasi - metrizable .
the space @xmath456 can be called the _ comb with its hand @xmath457 and teeth @xmath458 _ , @xmath459 ( compare with example iv.4.7 of @xcite ) . in
what follows , if @xmath9 and @xmath159 are gtses , the symbol @xmath460 denotes their @xmath100-product ( cf .
definition 4.6 of @xcite ) .
the space @xmath461 is @xmath212-metrizable by @xmath462 but @xmath463 is not @xmath212-quasi - metrizable .
similarly , @xmath464 is @xmath212-metrizable .
( see fact 4.10 in @xcite . )
the table of categories in @xcite , among other categories , says about the category @xmath465 of topological spaces , the category @xmath466 of bitopological spaces and about the category @xmath467 of bornological sets .
the categories @xmath100 , @xmath188 , @xmath468 of small generalized topological spaces and @xmath469 of locally small generalized topological spaces , as well as @xmath470 and @xmath471 , were introduced in @xcite and @xcite .
the categories @xmath473 and @xmath467 are all topological constructs ( cf .
@xcite , @xcite , @xcite,@xcite , @xcite and @xcite )
. since @xmath465 and @xmath467 are topological constructs , it is obvious that the category @xmath474 of bornological universes ( cf . remark 2.2.70 of @xcite ) is a topological construct , too .
let us define several more categories and answer the question whether they are topological constructs .
let @xmath475 be a boundedness in a set @xmath9 and let @xmath476 be a boundedness in a set @xmath159 .
we say that a mapping @xmath477 is @xmath478__-bounded _ _ ( in abbreviation : _ bounded _ ) if , for each @xmath479 , we have @xmath480 .
suppose that @xmath481 and @xmath482 are bornological biuniverses .
we say that a mapping @xmath477 is a _ bounded bicontinuous mapping _ from @xmath481 to @xmath482 if @xmath483 is bicontinuous with respect to @xmath484 and @xmath483 is @xmath478-bounded .
suppose that @xmath485 and @xmath486 are generalized bornological universes .
we say that a mapping @xmath487 is a _ bounded strictly continuous mapping _ from @xmath485 to @xmath486 if @xmath483 is both @xmath478-bounded and @xmath488-strictly continuous . 1 .
@xmath489 where objects are bornological biuniverses and morphisms are bounded bicontinuous mappings ; 2 .
@xmath490 where objects are generalized bornological universes and morphisms are bounded strictly continuous mappings ; 3 .
@xmath491 where objects are partially topological generalized bornological universes and morphisms are bounded strictly continuous mappings ; 4 .
@xmath492 where objects are small generalized bornological universes and morphisms are bounded strictly continuous mappings ; 5 .
@xmath493 where objects are partially topological small generalized bornological universes and morphisms are bounded strictly continuous mappings .
to check that , for instance , @xmath491 is a topological construct , we mimic the proof to theorem 4.4 of @xcite .
namely , let us consider a source @xmath494 of mappings @xmath495 indexed by a class @xmath496 where every @xmath497 is a partially topological generalized bornological universe and @xmath498 .
let @xmath96 be the @xmath100-initial generalized topology for @xmath499 in @xmath9 ( cf .
definition 4.3 of @xcite ) and let @xmath500 . for @xmath501 ,
let @xmath502 .
the canonical morphism @xmath503 is such that all mappings @xmath504 are morphisms in @xmath491 . for any object @xmath505 of @xmath491 and a mapping @xmath506
, we can observe that if all @xmath507 with @xmath508 are morphisms , then @xmath509 is a morphism of @xmath100 , so @xmath510 is a morphism of @xmath188 .
if all @xmath507 are bounded , then @xmath510 is bounded , too .
that @xmath511 and @xmath493 are topological can be proved by using similar arguments . | a concept of quasi - metrizability with respect to a bornology of a generalized topological space in the sense of delfs and knebusch is introduced .
quasi - metrization theorems for generalized bornological universes are deduced .
a uniform quasi - metrizability with respect to a bornology is studied .
the class of locally small spaces is considered and a possibly larger class of weakly locally small spaces is defined . the proofs and numerous examples are given in zf .
an example of a weakly locally small space which is not locally small is constructed under zf+cc .
several categories , relevant to generalized bornological universes , are defined and shown to be topological constructs . |
the search for novel quantum states in low - dimensional antiferromagnets ( af ) has triggered a significant activity in recent times @xcite .
a remarkable amount of theoretical studies has concerned the phase diagram of antiferromagnets where long - range magnetic order is suppressed by enhanced quantum fluctuations .
such a scenario can be established when the magnetic lattice dimensionality and the spin value are reduced or when the disorder is increased , either by means of heterovalent substitutions or by spin dilution @xcite .
further enhancement of quantum fluctuations can occur when antiferromagnetic interactions are competing as , for instance , in frustrated two - dimensional @xmath15 heisenberg af ( 2dqhaf ) on a square - lattice , with nearest neighbour ( @xmath9 ) and next - nearest neighbour ( @xmath10 ) antiferromagnetic couplings of the same order of magnitude @xcite .
recently , a prototype of frustrated 2dqhaf on a square - lattice has been discovered , the li@xmath3vosio@xmath6@xcite .
the analysis of the magnetic susceptibility , of the specific heat and of other quantities @xcite indicate that this compound is characterized by a ratio @xmath16 ranging from 1 to 4 @xcite .
the ground state is collinear @xcite , as theoretically expected @xcite , and the staggered magnetization reaches a value @xmath17v@xmath11 @xcite , consistent with the @xmath16 estimate @xcite .
this value is remarkably close to the one of a non - frustrated 2dqhaf @xcite and , therefore , at first sight one could think that , as far as @xmath16 is not close to @xmath18 and quantum fluctuations are not so strong , frustration does not affect sizeably the static magnetic properties of a 2dqhaf . in order to unravel the basic differences in the properties of frustrated and non - frustrated 2dqhaf
it is tempting to compare also the behaviour of these systems when disorder is introduced , for example , by spin dilution .
spin dilution has been widely investigated in prototypes of non - frustrated 2dqhaf , as la@xmath3cu@xmath4(zn , mg)@xmath5o@xmath6 @xcite , and evidence for the validity of the simple dilution model with a renormalization of the spin stiffness and for the disappearance of long - range magnetic order at the classical percolation threshold has been given @xcite .
at first , in a frustrated 2dqhaf on a square - lattice one would expect that the enhancement of quantum fluctuations leads to a suppression of the long - range order well below the percolation threshold .
however , one should also consider that , owing to the next - nearest neighbours coupling , the percolation threshold extends to much larger doping levels with respect to non - frustrated antiferromagnets @xcite .
hence , it is not trivial to say a priori how the magnetic properties vary upon increasing the spin dilution . in the following an experimental study of spin dilution effects in frustrated 2dqhaf on a square - lattice will be presented . in li@xmath3vosio@xmath6 spin dilution
can be achieved upon ti@xmath11 ( @xmath19 ) for v@xmath11 ( @xmath15 ) substitution .
it is found that spin dilution not only reduces the spin - stiffness by a factor @xmath12 , but also causes a reduction of the effective ratio @xmath13 .
the analysis of the sublattice magnetization curves shows that , at variance with non - frustrated 2dqhaf , spin dilution affects the low - temperature staggered magnetization only to a minor extent .
the @xmath14-dependence of the collinear ordering temperature and of other quantities will be discussed vis - - vis with the trend observed for non - frustrated 2dqhaf . in the following section
a description of the technical aspects involved in the sample preparation , magnetization , nmr and @xmath2sr measurements will be presented together with the experimental data .
then the results will be discussed in the light of models which turned out to be valid for non - frustrated spin - diluted 2dqhaf .
the final conclusions will be given in sect.iv .
powder samples of li@xmath3v@xmath4ti@xmath5osio@xmath6 were prepared in a platinum crucible by solid state reaction starting from a stoichiometric mixture of li@xmath3sio@xmath20 ( aldrich , 99.5% ) , tio@xmath3 ( aldrich , 99.8% ) and vo@xmath3 at 830 c , under vacuum , for 24h @xcite .
vo@xmath3 was prepared from a stoichiometric mixture of v@xmath3o@xmath21 and v@xmath3o@xmath20 by heating in a vacuum - sealed quartz tube at 650 c for 24h .
v@xmath3o@xmath20 itself was obtained by reducing v@xmath3o@xmath21 ( 99.9% , aldrich chem .
co. ) under hydrogen at 800 c. the sample purity was analyzed by means of x - ray powder diffraction and all peaks corresponded to the ones of li@xmath3vosio@xmath6 .
the lattice parameters of each sample were refined using the progamm celref .
the substitution of ti for v leads to no significative variation of the cell parameters , which is consistent with the rather similar ionic radii of the two ions ( @xmath22 and @xmath23 @xcite ) . upon varying @xmath14 from 0 to 0.15
the @xmath24 axes varies from 6.3683(4 ) to 6.3678(6 ) while the @xmath25 axes varies from 4.449(3 ) to 4.4502(4 ) .
the homogeneity of ti concentration in the different samples was verified by edx ( energy dispersive x - ray analysis ) .
magnetization measurements on li@xmath3v@xmath4ti@xmath5osio@xmath6 powders were performed using a quantum design xpms - xl7 squid magnetometer .
the temperature dependence of the susceptibility , defined as @xmath26 is shown in fig.[figsquid ] .
one observes a high temperature curie - weiss behaviour , a low - temperature maximum indicating the onset of antiferromagnetic correlations and a small kink in the 2 - 3 k range signaling a phase transition to a collinear ground - state . in the more doped sample , with 20% of ti
, a low - temperature upturn is noticed .
this upturn is typical of diluted 2dqhaf @xcite and originates from the correlated response of the spins around the @xmath19 impurity . above 20 k
the data can be conveniently fitted according to @xmath27 where @xmath28 is curie constant , @xmath29 the curie - weiss temperature and @xmath30 emu / mole @xcite the t - independent van - vleck susceptibility @xcite .
the x - dependence of the curie - weiss temperature is shown in fig.2 , together with the x - dependence of the temperature @xmath31 at which @xmath32 displays a maximum and of the transition temperature @xmath33 to the collinear phase .
this latter quantity was estimated from the peak in the derivative @xmath34 .
one notices basically a monotonous decrease of all three quantities with increasing dilution , as expected .
@xmath0li and @xmath1si nmr measurements have been carried out using standard radiofrequency pulse sequences . in particular ,
the nuclear spin - lattice relaxation rate @xmath35 was estimated from the recovery of the nuclear magnetization after a saturating pulse sequence . since in the powder sample at low - t @xmath0li nmr line
broadens due to the paramagnetic shift anisotropy , only a partial irradiation of the central and satellite lines is achieved .
this causes a departure of the recovery law from a single exponential and hence at low temperature , below about 3 k , it is more appropriate to fit the recovery of the nuclear magnetization with a stretched exponential form , namely @xmath36 .
the exponent @xmath37 decreased down to @xmath38 for @xmath39 k. on the other hand , the recovery law for @xmath1si was a single exponential over all the explored t - range .
the temperature dependence of @xmath0li and @xmath1si @xmath35 derived from the fit of the recovery laws following the aforementioned procedure are shown in figs . 3 and 4 , respectively .
one notices that while @xmath0li @xmath35 displays a peak at @xmath33 , which broadens upon increasing @xmath14 , @xmath1si @xmath35 does nt show any peak and the relaxation rate is found to decrease on cooling .
@xmath2sr measurements were performed on the emu instrument at the isis pulsed muon facility , using @xmath40 mev / c spin - polarized muons .
li@xmath3v@xmath4ti@xmath5osio@xmath6 powders were pressed on a silver sample - holder , whose background contribution to the muon asymmetry was determined from the slowly decaying oscillating signal in a 100 gauss transverse magnetic field , below @xmath33 .
in fact , below @xmath33 the external magnetic field sums up with the randomly oriented internal field @xmath41 and gives rise to a fast damping of the oscillating signal due to the muons stopping in the sample .
the background was estimated @xmath42 for @xmath43 and @xmath44 for @xmath45 .
these values were kept fixed for all the subsequent fits . by means of zero field ( zf ) muon spin relaxation measurements it is possible to extract the temperature dependence of the order parameter below @xmath33 , as previously done for the pure li@xmath3vosio@xmath6 @xcite .
below @xmath33 , superposed to an almost constant background , one observes a precessional signal at frequency @xmath46 ( fig.5 ) , with @xmath47 the muon gyromagnetic ratio and @xmath41 the local magnetic field at the muon generated by the collinear order . in a powder about 2/3 of the total signal oscillates while about 1/3 of the muons are in a longitudinal field configuration .
hence , one can write for the decay of the muon asymmetry @xmath48 with @xmath49 the asymmetry of the oscillating component and @xmath50 the one of the longitudinal component .
the ratio @xmath51 deviates from @xmath52 and is t - dependent since at a pulsed muon source the amplitude of the oscillating part is progressively filtered out as the precessional frequency increases .
the reduction in @xmath49 on cooling is actually slightly more pronounced than what one would expect , taking into account that at isis the width of the muon pulse is about 70 ns .
this suggests that a small part ( a fraction below 15% ) of the initial asymmetry is lost due to fast relaxing muons .
this observation is supported by preliminary experiments performed at psi facility .
there it was found that the asymmetry decay is characterized by a certain distribution of relaxation rates , typical of systems with impurities , and that a fraction of muons relax too fast to be detected at isis pulsed muon facility . in fig.[field ] the t - dependence of @xmath41 for @xmath43 is reported and compared to the one for @xmath8 @xcite .
one observes only a slight reduction of the saturation value of the local field at the muon and of @xmath33 , while the temperature dependence is unaffected .
above @xmath33 the decay of the muon is conveniently described by a nearly static kubo - toyabe function . as in the pure compound
the decay of the muon polarization can be conveniently fit using keren analytical approximation @xcite , which yields reliable results in the fast fluctuating limit , namely when @xmath53 , with @xmath54 the characteristic correlation time for the modulation of a magnetic field distribution of amplitude @xmath55 . at variance with li@xmath3vosio@xmath6 in the ti doped samples
one finds that @xmath56 .
however , keren approximation can still be used provided the fit is limited to a time @xmath57 @xcite . in fact
, the data above @xmath33 were accurately fit up to @xmath58 @xmath59 with the function ( fig.5 ) @xmath60 where the first term describes the spin - lattice relaxation driven by the fast fluctuations , while the second one is keren analytical approximation of kubo - toyabe function . by fixing @xmath61 from the high longitudinal field measurements above @xmath62 gauss and fitting @xmath2sr data at different magnetic fields ,
the t - dependence of @xmath54 was derived .
it was found that @xmath54 , at variance with the pure compound where it diverges on cooling , is @xmath63independent and more than an order of magnitude larger ( fig.[tauc ] ) .
in the weak doping limit , when the so called dilution model " should hold , the spin hamiltonian of a 2dqhaf with only n.n .
interactions can be written in the form @xmath64 where the @xmath8 hamiltonian is modified simply by taking into account the probability @xmath65 to find a spin at site @xmath66 .
this leads to a simple renormalization of the characteristic energy scales and the spin stiffness becomes @xcite @xmath67 .
one can try to extend this result to frustrated 2dqhaf on a square lattice by considering that also a second - nearest neighbour coupling is present . on qualitative grounds
one would expect that , since both the probability to find a pair of nearest - neighbour spins and of next - nearest - neighbours scale as @xmath68 , either @xmath9 or @xmath10 are renormalized by the same factor and hence , the effective degree of frustration @xmath16 is unaffected by dilution .
moreover , the spin - stiffness is expected to scale as @xmath68 as for the non - frustrated system .
it will be shown hereafter that although the spin - stiffness scales roughly as @xmath68 the experimental data evidence a modification of the effective @xmath16 ratio upon increasing @xmath14 .
preliminary information on the effects of spin - dilution can be obtained from the analysis of the doping dependence of the transition temperature @xmath69 , which shows a different trend with respect to non - frustrated 2dqhaf . in fig.[fig:4 ] @xmath69 , extracted either from susceptibility data or from the peak in @xmath0li @xmath35 , is shown .
the initial suppression rate of @xmath33 with ti doping @xmath70 , is not characterized by a @xmath28 value close to the one theoretically predicted and experimentally found for non - frustrated 2dqhaf , namely @xmath71 @xcite .
in fact , in li@xmath3v@xmath4ti@xmath5osio@xmath6 @xmath69 is rather well described by the expression @xmath72 ( see fig .
[ fig:4 ] ) , namely by @xmath73 .
this difference can be explained by resorting to the mean - field expression for @xmath69 @xcite @xmath74 where @xmath75 is the @xmath76 staggered magnetization , @xmath77 the in - plane correlation length in lattice units @xcite and @xmath78 the inter - layer coupling , which is reduced by a factor @xmath68 accounting for the probability to find two coupled spins in adjacent layers . from this expression
it is evident that the reduction of the staggered magnetization induced by dilution contributes to the reduction of @xmath69 .
however , at variance with non - frustrated compounds , in which a spin dilution of 5% was found to reduce the zero temperature magnetization already by about 9 % @xcite , in li@xmath3v@xmath4ti@xmath5osio@xmath6 the same amount of doping reduces it only by @xmath79 ( see fig .
[ field ] ) .
this explains why in li@xmath3v@xmath4ti@xmath5osio@xmath6 the effect of doping on @xmath69 is less pronounced .
moreover , it is observed that if in eq.[castroneto ] one neglects the reduction of the staggered magnetization with doping and one considers that @xmath80 , one finds an initial suppression of @xmath69 characterized by a coefficient @xmath81 , exactly the one experimentally observed for li@xmath3v@xmath4ti@xmath5osio@xmath6 . in the lower part of fig.2
the @xmath14-dependence of the spin - stiffness , estimated from eq.5 using the experimental values for @xmath69 and neglecting the reduction with @xmath14 of the staggered magnetization , is reported .
it is observed that the spin - stiffness is reduced by a factor @xmath82 by spin dilution .
a more accurate estimate of @xmath83 can be done taking into account the decrease of @xmath84 in eq.5 , derived from the reduction of the local field at the muon for @xmath85 ( fig.6 ) .
one finds that @xmath86 .
hence one concludes that the reduction of the spin - stiffness with @xmath14 is close to but not exactly the one that one would obtain by rescaling @xmath9 and @xmath10 by exactly the same factor @xmath68 . in this respect
it is interesting to analyze the effect of ti - doping on the effective ratio between the competing exchange couplings . a first evidence that @xmath13 is not @xmath87independent comes from the analysis of the susceptibility data ( see fig.[fig:4 ] ) .
the ratio between @xmath88 and the curie - weiss temperature @xmath29 is a measure of the degree of frustration @xcite .
in fact , while for @xmath89 this ratio is close to unity , on increasing @xmath10 it diminishes , reaching a minimum for @xmath90 and then increases again @xcite . in li@xmath3v@xmath4ti@xmath5osio@xmath6
this ratio is found to slightly decrease with @xmath14 from @xmath91 for @xmath8 to @xmath92 for @xmath45 . in principle
it is difficult to extract precise values of @xmath16 from the above ratios , however , if one considers that for small changes @xmath16 varies linearly with @xmath93 , the observed modifications imply a reduction in @xmath16 by 15 - 20 % for @xmath45 .
also the analysis of the temperature dependence of the nuclear spin - lattice relaxation rate @xmath35 suggests that the effective ratio @xmath16 is affected by spin - dilution .
@xmath35 can be written in terms of the components of the dynamical structure factor @xmath94 at the nuclear larmor frequency @xmath95 as @xmath96 where @xmath97 is the nuclear gyromagnetic ratio and @xmath98 is the form factor , which describes the hyperfine coupling of the spin excitations at wavevector * q * with the nuclei .
in non - frustrated two - dimensional @xmath15 antiferromagnets it was pointed out that the in - plane correlation length can be conveniently derived from the nuclear spin - lattice relaxation rate @xcite .
in fact , by applying scaling arguments for the amplitude and frequency of the collective spin excitations at wavevector * q * , one can write the nmr relaxation rate in terms of the correlation length @xmath99 as @xcite @xmath100 where @xmath101 , @xmath102 is the dynamical scaling exponent ( for @xmath103 @xmath104 @xcite ) , @xmath105 is a coefficient accounting for the reduction of the amplitude of spin excitations due to quantum fluctuations , @xmath106 is the heisenberg frequency @xcite , describing the uncorrelated spin fluctuation for @xmath107 and @xmath37 is a normalization factor which preserves the spin sum rule .
the form factors have been calculated on the basis of the hyperfine constants determined for the pure li@xmath3vosio@xmath6 @xcite . once @xmath102 is defined a one - to - one relationship between @xmath108 and @xmath99 is established , and one can determine @xmath99 from @xmath0li @xmath35 . then one can compare these results for @xmath99 with the recent theoretical predictions for the temperature dependence of @xmath99 , for different values of @xmath16 , based on the pure - quantum self - consistent harmonic approximation @xcite . in fig.[capriotti ]
@xmath99 extracted from @xmath0li nmr relaxation data in li@xmath3vosio@xmath6 are reported both in the assumptions that @xmath109 or that @xmath110 .
one observes that , above the crossover from 2d to 3d yielding the divergence of @xmath35 at @xmath33 , @xmath99 data derived for @xmath109 are in rather good agreement with the theoretical calculations for @xmath111 .
larger ratios of @xmath16 and a dynamical scaling exponent @xmath110 do not appear to be consistent with the experimental findings .
this suggests that the same scaling laws used for pure 2dqhaf could still be valid in the presence of frustration . for @xmath109 and @xmath112 ,
if the form factor is weakly @xmath113-dependent as for @xmath0li in li@xmath3vosio@xmath6 , one has that @xcite @xmath114 then , if @xmath9 and @xmath10 are reduced to the same extent by dilution and @xmath67 , the t - dependence of @xmath35 data should be @xmath14 independent once the temperature is rescaled to @xmath115 .
in fact , in diluted non - frustrated 2dqhaf as la@xmath3cu@xmath4zn@xmath5o@xmath6 , such a scaling does occur ( see fig.[fig:5 ] ) . on the other hand , in the diluted frustrated 2dqhaf li@xmath3v@xmath4ti@xmath5osio@xmath6
, such scaling does not seem to be accurate any longer , suggesting that also the effective ratio @xmath16 is changing .
in fact , one has to observe that while the spin hamiltonian for a diluted non - frustrated 2dqhaf on a square - lattice can still be mapped onto the same hamiltonian , provided the exchange coupling is renormalized , the spin hamiltonian for the frustrated 2dqhaf on a square lattice can no longer be mapped onto the same hamiltonian as dilution in the @xmath116 model yields the appearance of triangular spin configurations . after having presented these arguments suggesting a variation of the effective ratio @xmath13 and a negligible reduction of @xmath75 , the effect of spin vacancies on the distortion initially observed in undoped frustrated systems will be discussed @xcite .
as pointed out in the introduction , li@xmath3vosio@xmath6 is characterized by @xmath117 and hence by a two - fold degenerate collinear ground - state , with a magnetic wave vector which can be either * q*=(0 , @xmath118 ) or * q*=(@xmath118 , 0 ) . at low - t , after an ising transition , the spin system will eventually collapse in either one of the two collinear ground - states @xcite . now , in a real system a finite spin - lattice coupling exists and for @xmath117 it would favour a tetragonal to orthorombic distortion @xcite . in fact
, the behaviour of @xmath1si nmr spectra and @xmath35 observed in some li@xmath3vosio@xmath6 samples is consistent with such a distortion @xcite .
@xmath1si form factor @xmath119 is peaked at ( q@xmath120 , q@xmath121 ) and vanishes at the critical wavevector * q*=(0 , @xmath118 ) or * q*=(@xmath118 , 0 ) .
hence , on approaching @xmath33 , as the spectral weight shifts to the critical wavevector of the envisaged _ collinear _ order , @xmath35 should decrease and no peak observed .
however , in the pure li@xmath3vosio@xmath6 a peak in @xmath35 was observed suggesting that a distortion modifying @xmath1si form factor takes place @xcite . in ti - doped samples ( fig.4 )
no peak in @xmath1si @xmath35 is observed at @xmath33 , pointing out that ti impurities tend to hinder this distortion .
finally , we point out that above @xmath33 ti doping also hinders the very - low - frequency dynamics evidenced by @xmath2sr measurements in li@xmath3vosio@xmath6 ( fig .
[ tauc ] ) @xcite .
these dynamics were associated with the motion of domain walls separating regions were correlations with * q*=(0 , @xmath118 ) or * q*=(@xmath118 , 0 ) develop before the lattice distortion removes their degeneracy .
although a clear explanation of such a phenomenon goes beyond the aim of the present work , the observation of a t - independent much longer correlation time measured by @xmath2sr suggests that ti impurities might pin the motions of domain walls .
in conclusion , a throughout investigation of a spin diluted frustrated 2dqhaf system by means of nmr , @xmath2sr and susceptibility measurements has been presented .
it has been shown that the decrease of the magnetic ordering temperature is consistent with a reduction of the spin - stiffness by a factor @xmath12 and with a minor effect of spin dilution on the sublattice magnetization . the analysis of the magnetic susceptibility and of the nuclear spin - lattice relaxation rate shows that the effective ratio @xmath13 decreases with @xmath14 , namely that the two coupling constants are not renormalized in the same way by dilution .
this would actually mean that the spin hamiltonian of a frustrated 2dqhaf on a square - lattice can not be mapped onto the same square - lattice hamiltonian after spin dilution has occurred .
finally it was shown that the low frequency dynamics observed in pure li@xmath3vosio@xmath6 by means of @xmath2sr measurements disappears in li@xmath3v@xmath4ti@xmath5osio@xmath6 .
moreover the absence of a peak in @xmath1si nmr relaxation rate at the transition indicates that the distortion induced by frustration might be hindered by doping .
we would like to thank a. fubini for sending us the results of the theoretical calculations reported in ref . and
t.roscilde for useful discussions .
this work was supported by prin2004 national project `` strongly correlated electron systems with competing interactions '' .
see for example e. dagotto and t. m. rice , science 271 , 619 ( 1995 ) and e. dagotto , rep .
62 , 1525 ( 1999 ) and references therein s. chakravarty , b. i. halperin and d. r. nelson , phys . rev .
b 39 , 2344 ( 1989 ) a. rigamonti , f. borsa and p. carretta , rep .
61 , 1367 ( 1998 ) ; o. p. vajk , p. k. mang , m. greven , p. m. gehring and j. w. lynn , science 295 , 1691 ( 2002 ) p. chandra and b. doucot , phys . rev .
b 38 , 9335 ( 1988 ) r. melzi , p. carretta , a. lascialfari , m. mambrini , m. troyer , p. millet and f. mila phys .
85 , 1318 ( 2000 ) r. melzi , s. aldrovandi , f. tedoldi , p. carretta , p. millet and f. mila , phys . rev .
b 64 , 024409 ( 2001 ) p. carretta , n. papinutto , r. melzi , p. millet , s. gouthier , p. mendels and p. wzietek , j. phys . : condens .
matter 16 , s849-s856 ( 2004 ) a. bombardi , f. de bergevin , s. di matteo , l. paolasini , p. carretta , j. rodriguez - carvajal , p. millet and r. caciuffo , phys .
93 , 027202 ( 2004 ) p. chandra , p. coleman and a. i. larkin , phys .
64 , 88 ( 1990 ) h. j. schulz and t. a. l. ziman , europhys .
18 , 355 ( 1992 ) ; h. j. schulz , t. a. l. ziman and d. poilblanc , j. phys .
i ( france ) 6 , 675 ( 1996 ) b. keimer , n. belk , r. j. birgeneau , a. cassanho , c. y. chen , m. greven , m. a. kastner , a. aharony , y. endoh , r. w. erwin and g. shirane , phys .
b 46 , 14034 ( 1992 ) m. corti , a. rigamonti , f. tabak , p. carretta , f. licci and l. raffo , phys .
b 52 , 4226 ( 1995 ) p. carretta , a. rigamonti and r. sala , phys .
b 55 , 3734 ( 1997 ) yu - chang chen and a. h. castro neto , phys . rev .
b 61 , r3772 ( 2000 ) , a. l. chernyshev , y. c. chen , and a. h. castro neto phys .
b 65 , 104407 ( 2002 ) while for a two - dimensional square - lattice the percolation threshold for site - dilution is @xmath122 , preliminary estimates carried out by t. roscilde show that in presence of next - nearest neighbour couplings along the diagonals the percolation threshold increases to @xmath123 . p. millet and c. satto , mat .
33 , 1339 ( 1998 ) r.d .
shannon , acta cryst .
a 32 , 751 ( 1976 ) s. t. ting , p. pernambuco - wise , j. e. crow , e. manousakis and j. weaver , phys .
b 46 , 11772 ( 1992 ) p. carretta , r. melzi , n. papinutto and p. millet , phys .
88 , 047601 ( 2002 ) a. keren _ phys .
b _ * 50 * , 10039 ( 1994 ) see for example a. w. sandvik phys .
b 66 , 024418 ( 2002 ) and references therein a. r. mcgurn _ j. phys .
c _ * 12 * , 3523 ( 1979 ) s. bacci , e. gagliano and e. dagotto , phys . rev .
b 44 , 285 ( 1991 ) ; r. r. p. singh and r. narayanan , phys .
65 , 1072 ( 1990 ) ; g. misguich , b. bernu , and l. pierre phys . rev .
b 68 , 113409 ( 2003 ) a.rigamonti , p. carretta and n. papinutto in _ correlated spin dynamics and phase transitions in pure and in disordered 2d s=1/2 antiferromagnets from nmr - nqr _ , chapter in the book in honor of r. blinc ( springer - verlag 2005 ) l. capriotti , a. fubini , t. roscilde and v. tognetti , phys .
92 , 157202 ( 2004 ) c. weber , l. capriotti , g. misguich , f. becca , m. elhajal , and f. mila phys .
91 , 177202 ( 2003 ) f. becca and f. mila phys
89 , 037204 ( 2002 ) | @xmath0li and @xmath1si nmr , @xmath2sr and magnetization measurements in li@xmath3v@xmath4oti@xmath5sio@xmath6 , for @xmath7 , are presented .
the @xmath8 compound is a prototype of frustrated two - dimensional heisenberg antiferromagnet on a square - lattice with competing nearest ( @xmath9 ) and next - nearest ( @xmath10 ) neighbour exchange interactions .
ti@xmath11 ( s=0 ) for v@xmath11 ( s=1/2 ) substitution yields the spin dilution of the antiferromagnetic layers .
the analysis of the magnetization and of the nuclear spin - lattice relaxation rate shows that spin dilution not only reduces the spin - stiffness by a factor @xmath12 , but also causes the decrease of the effective ratio @xmath13 . moreover , the sublattice magnetization curves derived from zero - field @xmath2sr measurements in the collinear phase point out that , at variance with non - frustrated two - dimensional heisenberg antiferromagnets , spin dilution affects the low - temperature staggered magnetization only to a minor extent .
this observation is supported also by the @xmath14 dependence of the collinear ordering temperature .
the results obtained for the ti doped samples are discussed in the light of the results previously obtained in the pure @xmath8 compound and in non - frustrated two - dimensional heisenberg antiferromagnets with spin - dilution . |
given a positive integer @xmath5 , and a non - negative real @xmath4 , we consider a random geometric graph @xmath6 defined as follows .
the vertex set @xmath7 of @xmath8 is obtained by choosing @xmath5 points independently and uniformly at random in the square @xmath9 ^ 2 $ ] ( note that , with probability @xmath10 , no point in @xmath11 is chosen more than once , and thus we assume @xmath12 ) . for notational purposes , we identify each vertex @xmath13 with its corresponding geometric position @xmath14 , where @xmath15 and @xmath16 denote the usual @xmath17- and @xmath18-coordinates in @xmath11 . finally , the edge set of @xmath6 is constructed by connecting each pair of vertices @xmath19 and @xmath20 by an edge if and only if @xmath21 , where @xmath22 denotes the euclidean distance in @xmath11 .
random geometric graphs were first introduced in a slightly different setting by gilbert @xcite to model the communications between radio stations . since then , several closely related variants on these graphs have been widely used as a model for wireless communication , and have also been extensively studied from a mathematical point of view . the basic reference on random geometric graphs is the monograph by penrose @xcite .
the properties of @xmath23 are usually investigated from an asymptotic perspective , as @xmath5 grows to infinity and @xmath24 . throughout the paper ,
we use the following standard notation for the asymptotic behavior of sequences of non - negative numbers @xmath25 and @xmath26 : @xmath27 if @xmath28 ; @xmath29 if @xmath30 ; @xmath31 if @xmath27 and @xmath29 ; @xmath32 if @xmath33 . finally , a sequence of events @xmath34 holds _ asymptotically almost surely _ ( a.a.s . ) if @xmath35 .
it is well known that @xmath36 is a sharp threshold function for the connectivity of a random geometric graph ( see e.g. @xcite ) .
this means that for every @xmath37 , if @xmath38 , then @xmath23 is a.a.s .
disconnected , whilst if @xmath39 , then it is a.a.s .
connected . in order to ensure that we have a connected random geometric graph , we assume in the following that @xmath40 . given a connected graph @xmath8 , we define the _ graph distance _ between two vertices @xmath19 and @xmath20 , denoted by @xmath2 , as the number of edges on a shortest path from @xmath19 to @xmath20 .
observe first that any pair of vertices @xmath19 and @xmath20 must satisfy @xmath41 deterministically , since each edge of a geometric graph has length at most @xmath4 .
the goal of this paper is to provide upper and lower bounds that hold a.a.s . for the graph distance of two vertices in terms of their euclidean distance and in terms of @xmath4 ( see figure [ ap : fig : distxy2 ] ) .
_ related work . _
the particular problem has risen quite a bit of interest in recent years .
given any two @xmath42 , most of the work related to this problem has been devoted to study upper bounds on @xmath2 in terms of @xmath1 and @xmath4 , that hold a.a.s .
ellis , martin and yan @xcite showed that there exists some large constant @xmath43 such that for every @xmath44 , @xmath45 satisfies a.a.s .
@xmath46 for every @xmath19 and @xmath20 .
this result was extended by bradonjic et al .
@xcite for the range of @xmath4 for which @xmath23 has a giant component a.a.s .
, under the extra condition that @xmath47 .
friedrich , sauerwald and stauffer @xcite improved this last result by showing that the result holds a.a.s . for every @xmath19 and @xmath20 satisfying @xmath48 .
they also proved that if @xmath49 , a linear upper bound of @xmath2 in terms of @xmath1 is no longer possible .
in particular , a.a.s .
there exist vertices @xmath19 and @xmath20 with @xmath50 and @xmath51 .
the motivation for the study of this problem stems from the fact that these results provide upper bounds for the diameter of @xmath45 , denoted by @xmath52 , that hold a.a.s . , and
the runtime complexity of many algorithms can often be bounded from above in terms of the diameter of @xmath8 . for a concrete example
, we refer to the problem of broadcasting information ( see @xcite ) .
one of the important achievements of our paper is to show that @xmath53 a.a.s .
, provided that @xmath54 . by the result in @xcite , we know that such a result is false if @xmath49
. a similar problem has been studied by muthukrishnan and pandurangan @xcite .
they proposed a new technique to study several problems on random geometric graphs the so called _ bin - covering _ technique which tries to cover the endpoints of a path by bins .
they consider , among others , the problem of determining @xmath55 , which is the length of the shortest euclidean path connecting @xmath19 and @xmath20 .
recently , mehrabian and wormald @xcite studied a similar problem to the one in @xcite .
they deploy @xmath5 points uniformly in @xmath56 ^ 2 $ ] , and connect any pair of points with probability @xmath57 , independently of their distance .
mehrabian et al .
determine the ratio of @xmath55 and @xmath1 as a function of @xmath58 .
the following theorem is the main result of the paper .
[ ap : thm : main ] let @xmath6 be a random geometric graph with @xmath40 . a.a.s .
, for every pair of vertices @xmath59 we have : 1 . if @xmath60 , then @xmath61 , and 2 . if @xmath62 , then @xmath63 where @xmath64 in order to prove ( i ) , we first observe that all the short paths between two points must lie in a certain rectangle .
then we show that , by restricting the construction of the path on that rectangle , no very short path exists . for the proof of (
ii ) we proceed similarly .
we restrict our problem to finding a path contained in a narrow strip . in this case , we show that a relatively short path can be constructed .
we believe that the ideas in the proof can be easily extended to show the analogous result for @xmath65-dimensional random geometric graphs for all fixed @xmath66 .
\(1 ) we do not know if the condition @xmath60 in the lower bound in ( i ) can be improved .
( 2 ) the constant @xmath67 in the condition @xmath68 of ( ii ) ( as well as those in the definition of @xmath69 ) is not optimized , and could be made slightly smaller .
however , our method as it is , can not be extended all the way down to @xmath70 .
( 3 ) the error term in ( ii ) is @xmath71 which is @xmath72 iff @xmath73 .
hence , for @xmath54 , statement ( ii ) implies that a.a.s . @xmath74
thus improving the result in @xcite .
theorem [ ap : thm : main ] gives an upper bound on the diameter as a corollary .
first , observe that @xmath75 . from theorem 10 in @xcite for the particular case @xmath76
, one can easily deduce that if @xmath44 a.a.s .
@xmath77 moreover , note that a.a.s .
there exist @xmath19 and @xmath20 at distance @xmath78 ; the probability that the squares of side @xmath79 at the corners of @xmath11 contain no vertices is @xmath72 . applying theorem [ ap : thm : main ] to these vertices @xmath19 and @xmath20 , we obtain the following result . [
ap : cor : diam ] let @xmath6 be a random geometric graph with @xmath80 . a.a.s .
we have @xmath81 where @xmath69 is defined as in theorem [ ap : thm : main ] .
observe that our bound on the diameter stated in corollary [ ap : cor : diam ] improves the one in derived from @xcite whenever @xmath82 .
in order to simplify the proof of theorem [ ap : thm : main ] we will make use of a technique known as de - poissonization , which has many applications in geometric probability ( see @xcite for a detailed account of the subject ) . here
we sketch it .
consider the following related model of a random geometric graph given vertices @xmath19 and @xmath20 .
let @xmath83 , where @xmath84 is a set obtained as a homogeneous poisson point process of intensity @xmath10 in the square @xmath11 of area @xmath5 .
in other words , @xmath84 consists of @xmath85 points in the square @xmath86 chosen independently and uniformly at random , where @xmath85 is a poisson random variable of mean @xmath5 .
we add two labelled vertices @xmath19 and @xmath20 , whose position is also selected independently and uniformly at random in @xmath11 .
exactly as we did for the model @xmath23 , we connect by an edge @xmath19 and @xmath20 in @xmath7 if @xmath87 .
we denote this new model by @xmath88 .
the main advantage of defining @xmath89 as a poisson point process is motivated by the following two properties : the number of points of @xmath84 that lie in any region @xmath90 of area @xmath91 has a poisson distribution with mean @xmath91 ; and the number of points of @xmath84 in disjoint regions of @xmath11 are independently distributed . moreover , by conditioning @xmath88 upon the event @xmath92 , we recover the original distribution of @xmath23 . therefore , since @xmath93 , any event holding in @xmath88 with probability at least @xmath94 must hold in @xmath23 with probability at least @xmath95 .
we make use of this property throughout the article , and do all the analysis for a graph @xmath96 .
we will need the following concentration inequality for the sum of independently and identically distributed exponential random variables . for the sake of completeness we provide the proof here .
[ ap : lem : concentration ] let @xmath97 be independent exponential random variables and let @xmath98 . then , for every @xmath99 we have @xmath100 and for any @xmath101 we have @xmath102 by markov s inequality , we have for every @xmath103 @xmath104 where @xmath105 is the moment - generating function of an exponentially distributed random variable with parameter @xmath106 .
thus , @xmath107 setting @xmath108 , we have @xmath109 the lower tail is proved similarly .
our argument in this subsection depends only on the euclidean distance between @xmath19 and @xmath20 , but not on their particular position in @xmath11 .
thus , let @xmath110 and assume without loss of generality that @xmath111 and @xmath112 .
the next lemma shows that short paths between vertices are contained in small strips .
it is stated in the more general context of a geometric graph @xmath113 of radius @xmath4 , where the vertex set @xmath7 is a subset of points in the square @xmath11 ( not necessarily randomly placed ) , and edges connect ( as usual ) every pair of vertices at euclidean distance at most @xmath4 . for every @xmath114 ,
consider the rectangle @xmath115\times [ -\alpha,\alpha].$ ] [ ap : lem : strip ] let @xmath113 be a geometric graph with radius @xmath4 in @xmath11 , and let @xmath116 such that @xmath111 and @xmath112 .
suppose that @xmath117 , for some @xmath118 and @xmath119 .
then all paths of length at most @xmath120 from @xmath19 to @xmath20 are contained in @xmath121 .
suppose that there exists a path from @xmath19 to @xmath20 in at most @xmath120 steps .
let @xmath122 the vertex with largest @xmath18-coordinate in that path .
since @xmath123 $ ] , for any @xmath124 we have , @xmath125 therefore , @xmath126 where we used that @xmath127 . using that @xmath128
we have @xmath129 repeating the same argument for the vertex with smallest @xmath18-coordinate , we conclude that the path is contained in @xmath130\times [ -\alpha,\alpha]$ ] .
[ ap : prop : lb ] let @xmath131 be a random geometric graph on @xmath11 , with @xmath111 and @xmath112 . then , for every @xmath132 , we have that @xmath133 let @xmath134 and let @xmath135
. consider the event @xmath136 that all the paths from @xmath19 to @xmath20 of length @xmath120 are contained in the rectangle @xmath130\times [ -\alpha,\alpha]$ ] and let @xmath137 the event defined by condition .
if @xmath137 holds , then @xmath138 since @xmath119 , by lemma [ ap : lem : strip ] , @xmath139 .
denote by @xmath140 the vertex with largest @xmath17-coordinate inside the rectangle @xmath141\times [ -\alpha,\alpha]$ ] ( possibly @xmath142 if @xmath143 contains no other vertices of @xmath88 ) .
note that @xmath140 might not be connected to @xmath19 , but observe that its @xmath17-coordinate is always greater or equal to the @xmath17-coordinate of any vertex @xmath144 connected to @xmath19 ( see figure [ ap : fig : insider ] ) .
let @xmath145 be the @xmath17-coordinate of @xmath140 , and define the random variable @xmath146 . by definition , @xmath147 .
since @xmath96 , the number of vertices from @xmath7 inside a region of @xmath11 is a poisson random variable with mean equal to the area of that region .
hence , the random variable @xmath148 satisfies @xmath149 thus , @xmath148 is stochastically dominated by an exponentially distributed random variable @xmath150 of parameter @xmath151 .
we assume that @xmath148 and @xmath150 are coupled together in the same probability space , so that @xmath152 .
we proceed to define in a similar way the points @xmath153 and the values @xmath154 and @xmath155 , for any @xmath156 .
let @xmath153 be the vertex with largest @xmath17-coordinate inside the rectangle @xmath157\times [ -\alpha,\alpha]$ ] , and let @xmath154 be the @xmath17-coordinate of @xmath153 . define @xmath158 .
if @xmath159 contains no vertex of @xmath88 , then add an _ extra _ vertex @xmath160 ( so in that case @xmath161 and @xmath162 ) .
observe that @xmath163 , so the rectangles @xmath164 are disjoint .
moreover , @xmath165 for every @xmath166 ( by defining @xmath167 ) . therefore , @xmath168 are stochastically dominated by a sequence @xmath169 of i.i.d .
exponentially distributed random variables of parameter @xmath151 , such that @xmath170 for all @xmath166 . ,
@xmath140 coincide but @xmath171 and @xmath172 do not , scaledwidth=90.0% ] [ ap : fig : path ] note that the vertices @xmath173 may not induce a connected path in @xmath88 , since the euclidean distance between two consecutive ones may be greater than @xmath4 .
however , the fact that @xmath153 is the vertex with largest @xmath17-coordinate inside @xmath174\times[-\alpha,\alpha]$ ] and together with a straightforward induction argument yield to the following claim : if @xmath175 is a path contained in @xmath121 , then for every @xmath166 the @xmath17-coordinate of @xmath176 is at most @xmath154 ( see again figure [ ap : fig : insider ] ) .
we will now show that @xmath177 with the desired probability .
define @xmath178 expanding recursively from the relations @xmath179 and @xmath180 , we get @xmath181 let us consider the event that @xmath182 for all @xmath166 . in particular , this event implies that @xmath183 for all @xmath184 , and therefore @xmath185 . since each @xmath186
is exponentially distributed with parameter @xmath151 , @xmath187 so @xmath185 with probability at least @xmath188 .
moreover , notice that @xmath189 by lemma [ ap : lem : concentration ] , for all @xmath190 , @xmath191 where we have used that @xmath192 , for any @xmath193 . if @xmath194 , by definition of @xmath137 , @xmath195 and we are done .
thus , we may assume that @xmath196 .
for @xmath197 , from and , @xmath198 hence , if @xmath136 holds , then @xmath199 , and if @xmath200 holds , then @xmath201 .
thus , the proposition follows .
let @xmath88 be a random geometric graph in @xmath11 with labelled vertices @xmath19 and @xmath20 such that @xmath202 .
then we have @xmath203 with probability at most @xmath204 . as before , let @xmath110 and @xmath134 .
also let @xmath205 . since @xmath206 and @xmath207
, we have @xmath208 and @xmath209 by proposition [ ap : prop : lb ] , this implies that @xmath210 to finish the proof of statement ( i ) in theorem [ ap : thm : main ] , by de - poissonizing @xmath88 , we have that in @xmath23 , statement ( i ) in theorem [ ap : thm : main ] holds for our choice of @xmath19 and @xmath20 , with probability at least @xmath211 . note that this fact does not depend on the particular location of @xmath19 and @xmath20 in @xmath11 .
the statement follows by taking a union bound over all at most @xmath212 pairs of vertices . as in subsection
[ ap : ssec : i ] , we pick two points in @xmath11 , and put @xmath213 .
let @xmath69 be as in the statement of theorem [ ap : thm : main ] .
we assume first that @xmath111 and @xmath112 , and consider a poisson point process in the rectangle @xmath130\times[0,\alpha]$ ] , for a certain @xmath214 that will be made precise later .
let @xmath215 denote the random geometric graph on the rectangle @xmath121 , to which the points @xmath19 and @xmath20 are added .
we will show that the probability of having @xmath216 decays exponentially in @xmath217 .
for each point @xmath218 in @xmath121 with @xmath17-coordinate @xmath219 , define the rectangle @xmath220\times[0,\alpha ] , \qquad\text{where}\quad \rho = r-\frac{\alpha^2}{r } .\ ] ] we need the following auxiliary lemma .
[ ap : lem : rectangle ] for any vertex @xmath218 in @xmath121 , all vertices in @xmath221 are connected to @xmath218 ( see figure [ ap : fig : rectangle ] ) .
it is enough to show that the upper - left and the bottom - right corner of @xmath221 are at distance at most @xmath4 .
then all vertices inside @xmath221 are connected to one another , and in particular @xmath218 is connected to every vertex in @xmath221 .
a sufficient condition for that is @xmath222 or equivalently @xmath223 since @xmath224 for any @xmath193 , the lemma follows . ,
scaledwidth=40.0% ] [ ap : alternative ] let @xmath215 be a random geometric graph on @xmath121 , with @xmath111 and @xmath112 .
let @xmath225 and @xmath226 be constants and define @xmath227 .
then , for every @xmath228 , we have that @xmath229 set @xmath230 , and let @xmath137 be any positive constant satifying @xmath231 some elementary analysis shows that such @xmath137 must exist .
in fact , the equation @xmath232 has exactly two positive solutions @xmath233 and @xmath234 for any @xmath235 , and any @xmath236 satisfies .
let us consider the integer @xmath237 .
we will show that with very high probability there exists a path length at most @xmath120 between @xmath19 and @xmath20 .
such a path will only use vertices inside @xmath121 , but for technical reasons ( the last of the rectangles @xmath238 defined below might be further to the right than the point @xmath239 or possibly be outside of the square ) of the argument we extend the poisson point process of our probability space to the semi - infinite strip @xmath240 $ ] . we construct a sequence of vertices in a similar way as in the proof of proposition [ ap : prop : lb ] .
set @xmath241 , @xmath242 and @xmath167 .
we make the choice of @xmath243 for this subsection now more precise .
we set @xmath244 for some constant @xmath245 satisfying .
observe that the restriction @xmath246 implies that @xmath247 so our choice of @xmath243 is feasible , and moreover @xmath248 for each @xmath249 , define @xmath250\times[0,\alpha]$ ] , and let @xmath153 be the vertex with largest @xmath17-coordinate inside @xmath238 ( if @xmath238 is empty , then add an _ extra _ vertex @xmath160 ) .
define @xmath154 to be the @xmath17-coordinate of @xmath153 and @xmath251 . by the same considerations as in the proof of proposition [ ap : prop : lb ] ( but replacing @xmath159 by @xmath238 , @xmath4 by @xmath252 , and @xmath120 by @xmath253 ) , we deduce that @xmath254 are stochastically dominated by a sequence @xmath255 of i.i.d.exponentially distributed random variables of parameter @xmath243 , such that @xmath256 for all @xmath249 . moreover , since @xmath257 , we have @xmath258 thus , with probability at least @xmath259 , for every @xmath249 we have @xmath260 , and therefore @xmath183 .
this event implies that @xmath261 where @xmath262 , and also that we did not add any extra vertices ( i.e.all @xmath263 belong to the poisson point process in @xmath264 $ ] ) . by construction , each @xmath154 belongs to the rectangle @xmath265 for every @xmath266 .
hence , by lemma [ ap : lem : rectangle ] , the vertices @xmath267 form a connected path .
in view of all that , it suffices to show that @xmath268 with sufficiently large probability .
note that , if this event holds , then @xmath20 must belong to @xmath269 for some @xmath270 , and therefore @xmath271 is a connected path of length @xmath272 .
( observe that such a path is contained in @xmath121 , so our extension of the poisson point process to @xmath273 turned out to be harmless . )
recall that @xmath230 . using the upper - tail bound in lemma [ ap : lem : concentration ]
we obtain @xmath274 combining this together with , we infer that , with probability at least @xmath275 we have @xmath276 from the definition of @xmath120 , the range of @xmath217 and since @xmath277 , the event above implies @xmath278 \\ & \ge t \bigg[1 + \delta r^{-4/3 } \bigg(1 - ( f+1)\big(b^2+\frac{c}{b}\big ) \bigg ) \bigg ] > t,\end{aligned}\ ] ] so @xmath279 as desired . on the last step we used the fact that @xmath280 , which easily follows from .
this completes the proof of the proposition .
note that may be stronger than if we choose a constant @xmath137 which satisfies and maximises @xmath281 .
[ ap : prop : new_upperbound ] let @xmath69 as in the statement of theorem [ ap : thm : main ] , and let @xmath215 be a random geometric graph on @xmath121 , with @xmath111 and @xmath112 .
suppose that @xmath282 .
then we have @xmath283 with probability at most @xmath204 .
first , observe that , if @xmath284 , then @xmath285 , and the statement holds trivially .
thus , we assume henceforth that @xmath286 .
set @xmath287 , @xmath288 , @xmath289 , @xmath290 , @xmath291 and @xmath292 .
recall that @xmath293 we want to apply proposition [ ap : alternative ] with @xmath294 .
it is straightforward to check that the restrictions and @xmath226 , required in proposition [ ap : alternative ] hold .
we also need to show that @xmath295 .
notice that @xmath296 , since @xmath297 ; also @xmath298 , since @xmath299 ; and finally @xmath300 since @xmath301 .
moreover , @xmath302 .
note that this choice of constants combined with and implies @xmath303 the proof concludes by applying in the proof of proposition [ ap : alternative ] with this given @xmath217 , showing that the upper bound on @xmath304 is @xmath204 .
on the one hand , @xmath305 implies @xmath306 on the other hand , @xmath307 and @xmath308 imply @xmath309 where we have used that @xmath310 if @xmath311
. therefore , @xmath312 .
[ ap : cor : statement2 ] statement ( ii ) in theorem [ ap : thm : main ] is true .
observe that from the proof of proposition [ ap : alternative ] together with , @xmath313 with probability at least @xmath314 .
in particular , this event implies that @xmath140 is outside of the square @xmath315 \times [ 0,\alpha]$ ] .
moreover , also with probability @xmath314 , we can find some point @xmath316 in @xmath317\times[0,\alpha]$ ] .
it may happen that @xmath318 lies in @xmath319\times[0,\alpha]$ ] .
however , in that case , we can replace @xmath318 with @xmath316 , and therefore we found a @xmath19@xmath20 path of length @xmath272 with all internal vertices in @xmath320\times[0,\alpha]$ ] .
indeed , we will show now that we can always fit such a rectangle @xmath321\times[0,\alpha]$ ] , suitably rotated and translated , into the square .
we need first a few definitions .
consider two points @xmath322 and @xmath323 in @xmath324 . by symmetry
we may assume that @xmath325 and @xmath326 .
let @xmath327 be the angle of the vector @xmath328 with respect to the horizontal axis .
again by symmetry , we may consider @xmath329 $ ] .
we consider now two rectangles of dimensions @xmath330 placed on each side of the segment @xmath331 .
let @xmath332 be the rectangle to the left of @xmath328 , and let @xmath333 be the rectangle to the right of @xmath328 .
we will show that at least one of these rectangles contains a copy of @xmath334 fully contained in @xmath11 .
notice that the intersection of @xmath332 and @xmath333 with each of the halfplanes @xmath335 , @xmath336 , @xmath337 and @xmath338 gives @xmath339 triangles .
we call them @xmath340 , @xmath341 , @xmath342 and @xmath343 respectively .
all these triangles are right - angled , and denote by @xmath344 , @xmath345 , @xmath346 and @xmath347 the side of the corresponding triangle that it is parallel to the segment @xmath331 .
notice that @xmath348 and @xmath349 .
call a triangle @xmath350 , with @xmath351 and @xmath352 , _ safe _ if @xmath353 .
note that if @xmath340 and @xmath343 are safe or fully contained in the square , then @xmath332 contains the desired rectangle @xmath121 , and analogously for @xmath333 . since we assumed that @xmath354
, we have @xmath355 .
thus , @xmath340 and @xmath341 are safe .
if @xmath356 , that is @xmath357 , it is clear that either @xmath332 or @xmath333 contain the desired copy of @xmath334 .
thus , we may assume that @xmath358 .
we can also assume that both @xmath19 and @xmath20 are on the boundary of @xmath11 , as otherwise we extend the line segment @xmath331 to the boundary of the square , and the original rectangles are contained in the new ones .
recall that @xmath340 and @xmath341 are safe . if @xmath359 , then @xmath343 is completely contained in the square , and hence @xmath332 satisfies the conditions .
similarly , if @xmath360 , @xmath333 satisfies the conditions .
otherwise , @xmath361 , and the angle @xmath327 is at least @xmath362 , which contradicts our assumption on @xmath327 . again , by de - poissonizing @xmath88 , we can use proposition [ ap : prop : new_upperbound ] to show that for given @xmath19 and @xmath20 in @xmath6 , statement ( ii ) in theorem [ ap : thm : main ] holds with probability at least @xmath211 . by taking a union bound over all at most @xmath212 possible pairs of vertices , statement ( ii ) in theorem [ ap : thm : main ] follows
theorem [ ap : thm : main ] establishes a relation between the graph distance and the euclidean distance of two vertices @xmath19 and @xmath20 in @xmath23 that holds a.a.s .
simultaneously for all pairs of vertices . it would be interesting to find better concentration bounds on the values that @xmath2 can take with high probability .
also , we would like to characterize the probability distributions of @xmath363 and @xmath364 ( i.e. the expectation and variance of @xmath2 given @xmath1 ) .
what can we say about these distributions ? in the proof of statement ( ii ) in theorem
[ ap : thm : main ] , we define a new random variable that stochastically dominates @xmath2 and we give an upper bound for the probability that this random variable is too large .
this argument can be easily adapted in the case @xmath54 , and provide the upper bound @xmath365 .
similarly , the proof of statement ( i ) in theorem [ ap : thm : main ] can be adapted to give a lower bound on @xmath366 , but we need the further conditioning upon the event that @xmath1 is large enough .
m. bradonjic , r. elssser , t. friedrich , t. sauerwald and a. stauffer , efficient broadcast on random geometric graphs . , pp . 14121421 .
society for industrial and applied mathematics , 2010 .
r. b. ellis , j. l. martin and c. yan , random geometric graph diameter in the unit ball , 47(4 ) , pp .
421438 , 2007 .
gilbert , random plane networks , , 9(5 ) , pp .
533543 , 1961 .
a. goel , s. rai and b. krishnamachari , sharp thresholds for monotone properties in random geometric graphs , , 15 , pp .
364370 , 2005 . g. grimmett and d. stirzaker , , oxford u. p. , 2001 .
t. friedrich , t. sauerwald and a. stauffer , diameter and broadcast time of random geometric graphs in arbitrary dimensions , pp .
190199 , springer berlin heidelberg , 2011 .
a. merhabian and n. wormald , on the stretch factor of randomly embedded random graphs , http://arxiv.org/pdf/1205.6252v1.pdf , 2013 .
s. muthukrishnan and g. pandurangan , the bin - covering technique for thresholding random geometric graph properties , , pp . 989998 .
society for industrial and applied mathematics , 2005 .
m. penrose , the longest edge of the random minimal spanning tree , , 7(2 ) , pp . 340361 , 1997 .
m. penrose , , oxford studies in probability .
oxford u.p . | given any two vertices @xmath0 of a random geometric graph , denote by @xmath1 their euclidean distance and by @xmath2 their graph distance .
the problem of finding upper bounds on @xmath2 in terms of @xmath1 has received a lot of attention in the literature @xcite . in this paper , we improve these upper bounds for values of @xmath3 ( i.e. for @xmath4 above the connectivity threshold ) .
our result also improves the best known estimates on the diameter of random geometric graphs .
we also provide a lower bound on @xmath2 in terms of @xmath1 .
* keywords : * random geometric graphs , graph distance , euclidean distance , diameter . |
the nature of spin and/or charge inhomogeneities , especially in the form of stripes , in some cuprates and their involvement to high - temperature superconductivity are currently debate issues .
@xcite the stripe state is characterized by the self - organization of the charges and spins in the cuo@xmath7 planes in a peculiar manner , where the doped holes are arranged in one - dimensional ( 1d ) lines and form the so - called `` charge stripe '' separating the antiferromagnetic domains .
the stripe - ordered state minimizes the energy of the hole - doped antiferromagnetic system , thus leading to an inhomogeneous state of matter .
static one - dimensional charge and spin stripe order have been observed experimentally in a few special cuprate compounds , specifically in la@xmath8nd@xmath9sr@xmath10cuo@xmath11 @xcite and la@xmath12ba@xmath10cuo@xmath11 with @xmath13 .
@xcite similar signatures identified in la@xmath12sr@xmath10cuo@xmath11 ( lsco ) @xcite and other high temperature superconductors @xcite point to the possible existence of stripes , albeit of a dynamical or fluctuating nature .
a pivotal issue about this new electronic state of matter concerns whether it is compatible with superconductivity , and possibly even essential for the high transition temperatures , or it competes with the pairing correlations .
a prerequisite for addressing these issues is to understand the electronic structures of various stripe states in different cuprates , and to answer the question whether the stripe phase is intrinsically `` metallic '' or `` insulating '' , given its spin- and charge - ordered nature .
angle - resolved photoemission spectroscopy ( arpes ) study by zhou _
et al . _ in ( la@xmath14nd@xmath15sr@xmath16)cuo@xmath11 with static stripes have found the depletion of the low - energy excitation near the ( @xmath17 ) nodal region .
@xcite in another compound la@xmath18ba@xmath19cuo@xmath11 , a system where the superconductivity is heavily suppressed due to the development of the static spin and charge orders , valla _ et al .
_ have detected the high spectral intensity of the low - energy excitation in the vicinity of the ( @xmath17 ) nodal region [ while antinodal low - energy quasiparticle near @xmath20 are gapped ] .
@xcite the compound ( la@xmath21nd@xmath15sr@xmath10)cuo@xmath11 ( @xmath22 and @xmath23 ) with static one - dimensional stripe , seems to be an in - between system , in where the existence of spectral weight around the nodal region , though weak , has been identified .
@xcite meanwhile , optical conductivity measurements on the systems with a stripe phase also display the divergent results . in la@xmath24nd@xmath15sr@xmath19cuo@xmath11
@xcite and la@xmath18ba@xmath25sr@xmath10cuo@xmath11 , @xcite a finite frequency absorption peak with almost disappearance of the drude mode in the low - frequency conductivity in several experiments has been interpreted as collective excitations of charge stripes or as charge localization from the disorder created by nd or ba substitutions .
these observations may support the suggestion that such stripe - ordered state should be `` insulating '' in nature .
@xcite on the other hand , optical experiment on la@xmath18ba@xmath19cuo@xmath11 @xcite has observed a residual drude peak with a loss of the low - energy spectral weight below the temperature corresponding to the onset of charge stripe order , which indicates that stripes are compatible with the so - called nodal - metal state .
@xcite although , there have been some theoretical studies on the spectral and optical properties in the stripe phase in the past years , @xcite the contradictory observations in recent experiments as mentioned above have yet not been explained consistently in theoretical frame by adopting a realistic stripe model . in this paper , by using a stripe model in which the experimentally observed spin and charge structures at 1/8 doping are well reflected , we show that the spin domain - derived scattering will depress the zero - energy spectral weight around the nodal regions , while the charge domain - derived scattering will suppress mostly those around the antinodal regions and the hot spots . compared to the arpes data , this suggests that the different spectral weight distribution may result from the different relative strength of the spin and charge domain - derived scattering potentials inherently existing in these compounds .
meanwhile , a finite frequency peak in the optical conductivity appears with the disappearance of the drude peak in the case of the dominant spin domain - derived scattering . while , when the charge domain - derived scattering is comparable to the spin one , a residual drude peak exists with the disappearance of the finite energy peak
this suggests that both the `` insulating '' and `` metallic '' properties are intrinsic to the stripe state without introducing another distinct metallic phase .
the rest of this paper is organized as follows . in sec .
ii , we introduce the model hamiltonian and carry out the analytical calculations . in sec .
iii , we present the numerical calculations and discuss the results . in sec .
iv , we present the conclusion .
as the above discussed compounds have a doping density at or near 1/8 , we will in this paper consider the 1/8 doping antiphase vertical stripe state . a schematic illustration of its charge and spin pattern is presented in fig .
1 . the charge stripes , with a unit cell of 8 lattice sites ( note for 1/8 doping , there is one hole for every two sites along the length of a charge stripe ) , act as antiphase domain walls for the magnetic order , so that the magnetic unit cell is twice as long as that for charge order . due to the periodical modulation of the stripe order ,
the electrons moving in the state will be scattered by the modulation potentials .
after fourier transformation , the potential @xmath26 can be written as the scattering term between the state @xmath27 and those at @xmath28 with @xmath29 .
following ref . , we expect that the terms @xmath30 and @xmath31 will be the dominant spin and charge domain - derived scattering term , and will be relabeled as @xmath0 and @xmath1 in the following , respectively .
the weaker higher harmonic terms will be neglected here . in the coexistence with the superconducting ( sc ) order ,
the model hamiltonian can be written as a @xmath32 matrix for @xmath27 in the reduced brillouin zone , @xmath33 where , the prime denotes the summation over the reduced brillouin zone .
@xmath34 is a column vector with its elements @xmath35 for @xmath36 , and @xmath37 for @xmath38 .
both @xmath39 and @xmath40 are @xmath41 matrix with @xmath42 and @xmath43 as for the tight - binding energy band , we will choose the following form , @xcite @xmath44 where , @xmath45 is the doping density , and a @xmath46-wave sc order parameter @xmath47 is assumed .
generally , the charge modulation will induce the modulation of the sc order leading to the finite momentum pairs .
however , in the present study , one of our aim is to examine the effect of the spin ( charge ) domain - derived scattering on the sc order .
in this regard , the average value of the sc order parameter is relevant and the modulation of the sc order will be ignored .
we have checked the effect of this modulation and found no qualitative change in the results presented in fig .
2 . in the following
, @xmath48 is taken as the energy unit , @xmath49 , @xmath50 , @xmath51 .
this dispersion can be derived from the slave - boson mean - field calculation of the @xmath52 model @xcite , and in this way the parameters @xmath53 , @xmath54 and @xmath55 are determined self - consistently .
here we take it as a phenomenological form . in a self - consistent calculation ,
the hamiltonian is first diagonalized by a unitary matrix @xmath56 with a set of trial values of @xmath53 , @xmath54 and @xmath55 for given potentials @xmath0 and @xmath1 .
then @xmath53 , @xmath54 and @xmath55 are self - consistently calculated by using the relations : @xmath57 ( to get the @xmath46-wave pairing , the sign before @xmath53 takes @xmath58 for @xmath59 and @xmath60 for @xmath61 , where @xmath62 and @xmath63 denote the unit vectors along @xmath64 and @xmath65 directions , respectively . ) , @xmath66 , and @xmath67 , respectively .
reformularization of the expressions of @xmath53 , @xmath54 and @xmath55 in terms of eigenfunctions and eigenvalues of the hamiltonian , one obtains the self - consistency relations @xmath68 \nonumber \\
\chi_{0}&=&\frac{1}{n}\sum_{k}(\cos k_{x}+\cos k_{y})\sum^{16}_{m=1}u_{1m}(k)u^{\dag}_{m1}(k)f[e_{m}(k ) ] \nonumber \\ n&=&\frac{2}{n}\sum_{k}\sum^{16}_{m=1}u_{1m}(k)u^{\dag}_{m1}(k)f[e_{m}(k)],\end{aligned}\ ] ] where , @xmath69 is the eigenvalue of the hamiltonian , @xmath70 the elements of the matrix @xmath56 , and @xmath71 $ ] is the fermi - dirac distribution function .
then , the single particle green functions @xmath72 can be expressed as @xmath73 and the spectral functions is @xmath74
we first present in fig .
2 the self - consistent results of the sc order parameter as a function of @xmath0 and @xmath1 . while the scattering from both spin and charge domain - derived scattering potentials in the stripe state leads to the suppression of the sc order parameter , the charge domains are more compatible with superconductivity than spin domains , as can be seen from fig .
this may support the statement that the sc pairing in the stripe state occurs most strongly within the charge stripes .
@xcite on the other hand , an interesting feature is that the sc order parameter will be zero at the spin domain - derived scattering potential @xmath75 in the absence of the charge domain - derived scattering , however , it will develop a noticeable value after turning on the charge domain - derived scattering potential , as shown in fig .
this shows that the charge domain - derived scattering will lead to the emergency of the sc order which is otherwise destroyed by the spin only scattering . in fig .
3 , we present the distribution of the low - energy spectral weight in the original brillouin zone ( integrated over an energy window @xmath76 about @xmath77 ) in the 1/8 antiphase stripe state for different spin ( charge ) domain - derived scattering potential @xmath0 ( @xmath1 ) .
let us first look at the limit where only the spin domain - derived scattering is included , i.e. , @xmath78 with @xmath79 , one will find that the spectral weight around the nodal region is suppressed heavily[see fig .
3(a ) ] . at another limit where only the charge domain - derived scattering is included ( @xmath80 with @xmath81 ) ,
the spectral weight around the nodal region is recovered and those around the hot spot ( the cross of the fermi surface with the line @xmath82 ) and near the antinodal region are suppressed[see fig .
3(b ) ] . starting from the limit of @xmath78 and fixing @xmath79 , the spectral weight will redistribute gradually from the antinodal region to the nodal region with the increase of the charge domain - derived scattering potential @xmath1 , as shown in figs
. 3(c ) and ( d ) .
when two scattering potentials are comparable , the strongest spectral weight situates around the nodal region , and at the meantime noticeable spectral weights along the whole fermi surface is presented .
therefore , the divergent features observed in arpes measurements by zhou _
et al . _ in ( la@xmath14nd@xmath15sr@xmath83)cuo@xmath11 @xcite in which the low - energy excitations near the nodal region are depleted , and by valla _
_ in la@xmath18ba@xmath19cuo@xmath11 @xcite in which the high spectral intensity of the low - energy excitation in the vicinity of the nodal region is detected are consistently reproduced here by a change of the relative strength between the charge and spin domain - derived scatterings .
this consistent accounting enables us to propose that the spin domain - derived scattering dominates over the charge one in the former system while the scattering strengthes of them are comparable in the latter system . in the presence of the spin ( charge ) domain - derived potential ,
quasiparticles near the fermi surface will be scattered from @xmath84 to @xmath85 ( n=1 for the spin domain - derived potential , n=2 for the charge one ) , for the 1/8 antiphase vertical stripe configuration shown as fig .
1 . this gives rise to two scattering channels from the spin domain with potential @xmath0 , @xmath86 and two scattering channels from the charge domain with potential @xmath1 , @xmath87 strong potential scattering will destruct those parts of the fermi surface connected by the above mentioned scattering wave vectors .
because the scattering wave vectors @xmath88 and @xmath89 are close to the transferred momenta from the node to node scattering , so it will lead to a depletion of the spectral weight near the nodal region as shown in fig .
3(a ) . on the other hand , the scattering wave vectors
@xmath90 and @xmath91 , which is near the connecting wave vectors between the two approximately parallel segments of the fermi surface near the antinodal and hot spot region , the scatterings with these wave vectors will suppress the spectral weights around the antinodal and hot spot regions [ fig .
3(b ) ] . now
, we turn to the discussion of the in - plane optical properties in the 1/8 antiphase stripe state , and to see how they are influenced by the scattering from the spin and charge domains .
we will fix the temperature at @xmath92 in all calculations , in order to avoid the influence from the temperature induced change in the scattering rate .
we consider an electric field applied in the @xmath64 direction , which is perpendicular to the stripe . from the kubo formula for the optical conductivity , the real part of the optical conductivity is @xmath93/\omega$ ] .
the imaginary part of the current - current correlation function im@xmath94 $ ] is given by @xmath95=&&\frac{\pi}{n}\sum_{k}{'}\sum^{16}_{j , l=1}v^{jj}(k)v^{ll}(k ) \nonumber \\ & & \times \int d\omega^{'}[f(\omega+\omega^{'})-f(\omega^ { ' } ) ] \nonumber \\ & & \times a_{jl}(k,\omega^{'})a_{lj}(k,\omega+\omega^{'}).\end{aligned}\ ] ] here , @xmath96 is the diagonal element of the quasiparticle group velocity in the matrix form @xmath97 figs .
4(a)-4(d ) show the results for the optical conductivity calculated with the same scattering potentials as used to get fig .
3(a)-3(d ) . with only spin domain - derived scattering[fig .
4(a ) ] , no drude - like component appears at zero frequency in the optical conductivity , instead a finite frequency conductivity peak occurs around 0.3 .
this indicates that the system exhibits the `` insulating '' property .
@xcite when only charge domain - derived scattering is considered[fig .
4(b ) ] , the drude - like peak shows up and at the meantime the finite frequency peak remains .
optical conductivity involves the contribution from the quasiparticle excitations along the whole fermi surface weighted by the quasiparticle group velocity . due to the relative flat band structure near the antinodal region for the high-@xmath6 cuprates ,
the zero frequency optical conductivity mainly comes from the quasiparticle excitations around the nodal region . in the case of
only spin domain - derived scattering , the nodal region of the fermi surface is gapped and therefore the quasiparticle spectral weight is suppressed around the nodal region as shown in fig .
3(a ) , so that the zero - frequency drude - like peak is absent and a finite frequency peak with its position being equal to the gap ( @xmath98 ) occurs .
for the charge domain - derived scattering , the gap opens around the hot spots and near the antinodal , but a large spectral weight situates around the nodal region , as can be seen from fig .
thus , the drude - like peak emerges and the finite frequency peak remains ( it is now situates at @xmath99 ) . as shown in fig .
3(c ) , with the increase of the charge domain - derived scattering @xmath1 , the gap near the nodal region which is resulted from the spin domain - derived scattering will be suppressed gradually and correspondingly the spectral weight will be enhanced . as a result , the finite frequency peak in the optical conductivity
is shifted to lower frequency and the zero frequency component is lifted up gradually[fig .
when the charge domain - derived scattering is comparable to the spin one , the quasiparticles have noticeable spectra weight along the entire fermi surface with its largest weight around the nodal region[fig .
3(d ) ] , then the drude - like mode occurs at the zero frequency , and the finite frequency peak fades away and merges into the drude - like peak , as shown in fig .
the calculated results for the optical conductivity presented in figs .
4(c ) and 4(d ) are consistent well with the experimental observations in the stripe state of la@xmath24nd@xmath15sr@xmath19cuo@xmath11 @xcite and la@xmath18ba@xmath19cuo@xmath11 @xcite , respectively .
we now discuss the implication of our theoretical results .
as noted in the introduction , in la@xmath24nd@xmath15sr@xmath19cuo@xmath11 system , arpes experiment has found that there is little or no low - energy spectral weight near the nodal region , @xcite and optical conductivity experiment has observed a finite frequency peak with almost the disappearance of the drude mode , indicating an `` insulating '' stripe state . @xcite these spectroscopic features can be reproduced here with a strong spin domain - derived scattering potential @xmath79 and a weak charge domain - derived potential @xmath100 and @xmath78 , as shown in figs .
3(a ) , 3(c ) , 4(a ) and 4(c ) . interestingly , in this parameter regime for the spin and charge domain - derived scattering , the sc order is destroyed as can be seen from fig .
this is in consistent with the experimental fact that la@xmath24nd@xmath15sr@xmath19cuo@xmath11 is nonsuperconducting . in another cuprate la@xmath18ba@xmath19cuo@xmath11
, arpes spectra have identified the existence of high spectral intensity around the nodal region , @xcite and the optical conductivity measurement has observed a residual drude peak without the finite frequency peak , @xcite pointing to a so - called nodal metal state .
@xcite when comparable spin and charge domain - derived scattering potentials are assumed such as @xmath79 and @xmath80 , we can reproduce these features consistently , as shown in figs . 3(d ) and 4(d ) . on the other hand ,
a weak superconductivity emerges in the otherwise nonsuperconducting regime ( when only spin scattering potential @xmath101 is considered ) with the increase of the charge domain - derived scattering potential[see fig .
this suggests that the weak superconductivity in la@xmath18ba@xmath19cuo@xmath11 is likely beneficial from the metallic behaviors of the stripe state originated from a sufficient charge domain - derived scattering .
the above mentioned consistent accounting for both divergent spectroscopic features observed in two families of high-@xmath6 cuprates indicates that the stripe state may be intrinsically `` insulating '' or `` metallic '' , depending on the relative strength of the spin and charge domain - derived scattering potentials . specifically , a large spin domain - derived scattering potential favors the `` insulating '' state , while a large charge domain - derived scattering potential the `` metallic '' state .
we have calculated the sc order parameter , the spectral function and the optical conductivity in a stripe model with spin and charge domain - derived scattering potentials ( @xmath0 and @xmath1 ) .
the self - consistent calculation of the sc order parameter shows that the charge domain - derived scattering is less effective than the spin scattering on the suppression of superconductivity , and may even lead to the emergency of the sc order which is otherwise destroyed by the spin only scattering . for @xmath2 , the zero - energy spectral weight disappears around the nodal points , and a finite energy peak appears in the optical conductivity with almost the disappearance of the drude peak .
but for @xmath4 , the spectral weight concentrates on the nodal region , and a residual drude peak exists in the optical conductivity without the finite energy peak
. these results consistently account for the divergent spectroscopic properties observed experimentally in two families of high-@xmath6 cuprates , and demonstrate that both the `` insulating '' and `` metallic '' behavior may be the intrinsic properties of the stripe state , depending on the relative strength of the spin and charge domain - derived scattering potentials .
this project was supported by national natural science foundation of china ( grant no .
10525415 ) , the ministry of science and technology of science ( grants nos . 2006cb601002 , 2006cb921800 ) , the china postdoctoral science foundation ( grant no .
20080441039 ) , and the jiangsu planned projects for postdoctoral research funds ( grant no .
0801008c ) .
55 s. a. kivelson and i. p. bindloss , e. fradkin , v. oganesyan , j. m. tranquada , a. kapitulnik , and c. howald , rev .
phys . * 75 * , 1201 ( 2003 ) . j. m. tranquada , b. j. sternlieb , j. d. axe , y. nakamura , and s. uchida , nature * 375 * , 561 ( 1995 ) .
x. j. zhou , p. bogdanov , s. a. kellar , t. noda , h. eisaki , s. uchida , z. hussain , and z .- x .
shen , science * 286 * , 268 ( 1999 ) .
j. m. tranquada , h. woo , t. g. perring , h. goka , g. d. gu , g. xu , m. fujita , and k. yamada , nature * 429 * 534 ( 2004 ) .
p. abbamonte , a. rusydi , s. smadici , g. d. gu , g. a. sawatzky and d. l. feng , nat . phys . * 1 * 155 ( 2005 ) .
cheong , g. aeppli , t. e. mason , h. mook , s. m. hayden , p. c. canfield , z. fisk , k. n. clausen , and j. l. martinez , phys .
67 * , 1791 ( 1991 ) .
t. e. mason , g. aeppli , and h. a. mook , phys .
lett . * 68 * , 1414 ( 1992 ) .
a. bianconi , n. l. saini , a. lanzara , m. missori , t. rossetti , h. oyanagi , h. yamaguchi , k. oka , and t. ito , phys .
lett . * 76 * , 3412 ( 1996 ) .
k. yamada , c. h. lee , k. kurahashi , j. wada , s. wakimoto , s. ueki , h. kimura , y. endoh , s. hosoya , g. shirane , r. j. birgeneau , m. greven , m. a. kastner , and y. j. kim , phys .
b * 57 * , 6165 ( 1998 ) . b. o. wells , y. s. lee , m. a. kastner , r. j. christianson , r. j. birgeneau , k. yamada , y. endoh , and g. shirane , science * 277 * , 1067 ( 1997 ) . y. s. lee , r. j. birgeneau , m. a. kastner , y. endoh , s. wakimoto , k. yamada , r. w. erwin , s .- h . lee , and g. shirane , phys .
b * 60 * , 3643 ( 1999 ) .
h. a. mook , p. dai , s. m. hayden , g. aeppli , t. g. perring , and f. doan , nature * 395 * , 580 ( 1998 ) .
t. valla , a. v. fedorov , j. lee , j. c. davis , and g. d. gu , science * 314 * , 1914 ( 2006 ) .
x. j. zhou , t. yoshida , s. a. kellar , p. v. bogdanov , e. d. lu , a. lanzara , m. nakamura , t. noda , t. kakeshita , h. eisaki , s. uchida , a. fujimori , z. hussain , and z .- x .
shen , phys .
86 * , 5578 ( 2001 ) .
m. dumm , and d. n. basov , seiki komiya , yasushi abe , and yoichi ando , phys .
lett . * 88 * , 147003 ( 2002 ) .
m. ortolani , p. calvani , s. lupi , u. schade , a. perla , m. fujita , and k. yamada , phys .
b * 73 * , 184508 ( 2006 ) . c. castellani , c. di castro , m. grilli , a. perali , physica c * 341 - 348 * , 1739 ( 2000 ) . c. c. homes , s. v. dordevic , g. d. gu , q. li , t. valla , and j. m. tranquada , phys . rev . lett . * 96 * , 257002 ( 2006 ) . y. ando , a. n. lavrov , s. komiya , k. segawa , and x. f. sun , phys .
lett . * 87 * , 017001 ( 2001 ) .
x. j. zhou , t. yoshida , a. lanzara , p. v. bogdanov , s. a. kellar , k. m. shen , w. l. yang , f. ronning , t. sasagawa , t. kakeshita , t. noda , h. eisaki , s. uchida , c. t. lin , f. zhou , j. w. xiong , w. x. ti , z. x. zhao , a. fujimori , z. hussain , and z .- x .
shen , nature ( london ) * 423 * , 398 ( 2003 ) .
m. dumm , s. komiya , y. ando , and d. n. basov , phys .
lett . * 91 * , 077004 ( 2003 ) .
m. sutherland , d. g. hawthorn , r. w. hill , f. ronning , s. wakimoto , h. zhang , c. proust , e. boaknin , c. lupien , l. taillefer , r. liang , d. a. bonn , w. n. hardy , r. gagnon , n. e. hussey , t. kimura , m. nohara , and h. takagi , phys . rev .
b * 67 * , 174520 ( 2003 ) .
y. s. lee , k. segawa , y. ando , and d. n. basov , phys .
b * 70 * , 014518 ( 2004 ) .
t. tohyama , s. nagai , y. shibata , and s. maekawa , phys .
rev . * 82 * , 4910 ( 1999 ) .
r. s. markiewicz , phys .
b * 62 * , 1252 ( 2000 ) .
i. martin , g. ortiz , a. v. balatsky , and a. r. bishop , europhys .
56 * , 849 ( 2001 ) .
j. lorenzana , and g. seibold , phys .
* 90 * , 066404 ( 2003 ) .
a. j. millis , and m. r. norman , phys .
b * 76 * , 220503(r ) ( 2007 ) .
p. a. lee , n. nagaosa , and x. g. wen , rev .
* 78 * , 17 ( 2006 ) .
j. x. li , c. y. mou , and t. k. lee , phys . rev .
b * 62 * , 640 ( 2000 ) .
e. berg , e. fradkin , e .- a .
kim , s. a. kivelson , v. oganesyan , j. m. tranquada , and s. c. zhang , phys .
* 99 * , 127003 ( 2007 ) .
the term `` insulating '' state used here follows refs . to indicate a strong suppression of drude peak in the optical conductivity
in fact , the spectral weight at the fermi level will not be fully gapped out , so it is not a true insulating state .
we use the term here is to facilitate our comparison with the experiments[refs . ] . | we investigate the superconducting order parameter , the spectral and optical properties in a stripe model with spin ( charge ) domain - derived scattering potential @xmath0 ( @xmath1 ) .
we show that the charge domain - derived scattering is less effective than the spin scattering on the suppression of superconductivity . for @xmath2 ,
the spectral weight concentrates on the ( @xmath3 ) antinodal region , and a finite energy peak appears in the optical conductivity with the disappearance of the drude peak .
but for @xmath4 , the spectral weight concentrates on the ( @xmath5 ) nodal region , and a residual drude peak exists in the optical conductivity without the finite energy peak
. these results consistently account for the divergent observations in the arpes and optical conductivity experiments in several high-@xmath6 cuprates , and suggest that the `` insulating '' and `` metallic '' properties are intrinsic to the stripe state , depending on the relative strength of the spin and charge domain - derived scattering potentials . |
the liquid - vapor fractionation ratio , @xmath0 , is defined as @xmath1 where @xmath2 is the mole fraction of isotope @xmath3 , @xmath4 denotes the liquid phase , and @xmath5 denotes the vapor phase . in the second equality ,
@xmath6 is the helmholtz free energy corresponding to the process @xmath7 in this work we consider the dilute deuterium ( d ) limit which reflects the situation found in the earth s atmosphere where it is 6000 times less common than h. in this limit , we consider the free energy of exchanging a single d atom in a vapor water molecule with an h atom in a liquid water molecule , with all other molecules being h@xmath8o .
the free energy difference can be calculated from the thermodynamic integration expression @xcite @xmath9 where @xmath10 and @xmath11 are the kinetic energy expectation values for a hydrogen isotope of mass @xmath12 in a water molecule ho@xmath3 in the liquid and vapor phases respectively .
the kinetic energy can be calculated exactly for a given potential energy model of water using a path integral molecular dynamics ( pimd ) simulation .
these simulations exploit the exact isomorphism between a system of quantum mechanical particles and that of a set of classical ring polymers in which the spread of a polymer is directly related to that quantum particle s position uncertainty@xcite .
the kinetic energy for the particle @xmath3 in the molecule ho@xmath3 can be calculated from these simulations using the centroid virial estimator @xcite @xmath13 where t is the temperature and k@xmath14 is the boltzmann constant . here ,
@xmath15 is vector from the @xmath16th bead representing particle @xmath3 to the center of the ring polymer and @xmath17 is the force on that bead as shown in fig .
[ fig : waterbeads ] .
the first term in this expression is the classical kinetic energy and is independent of the surrounding environment , thus it is identical in the vapor and liquid phase .
the second term is the kinetic energy associated with confinement of a quantum particle .
this confinement depends on the forces exerted on the particle by the surrounding molecules .
equilibrium fractionation is thus an entirely quantum mechanical phenomenon .
pimd simulations were performed using 1000 water molecules with @xmath18=32 ring polymer beads used for the imaginary time discretization . previously described evolution and thermostatting procedures were used @xcite .
the computational cost of these calculations was reduced by using the ring polymer contraction technique with a cut - off of 5 , which for this system size leads to more than an order of magnitude speed - up compared to a standard pimd implementation @xcite . the integral in eq .
[ eq : deltaa ] was performed using the midpoint rule with 11 masses @xmath12 evenly spaced between @xmath19 and @xmath20 .
calculations were performed at the experimental coexistence densities@xcite . to model the interactions within and between water molecules
, we used the q - spcfw @xcite and q - tip4p / f @xcite models which have previously been shown to accurately reproduce many of water s properties in pimd simulations of liquid water .
both models are flexible , use point charges , and have a harmonic description of the bending mode .
however , while q - spcfw uses a purely harmonic description of the oh stretch @xmath21 q - tip4p / f contains anharmonicity by modeling the stretch as a morse expansion truncated at fourth order @xmath22 .
\label{eq : aharm}\ ] ] here @xmath23 is the distance between the oxygen and hydrogen atom and the parameters are given in refs .
@xcite and @xcite . the anharmonicity in the q - tip4p / f model makes the observed quantum mechanical effects much smaller than previously predicted from harmonic or rigid models and gave rise to the idea of `` competing quantum effects '' in water@xcite . both models have previously been shown to accurately reproduce many of water s properties in pimd simulations of liquid water . due to their simple potential form such models are generally less transferable to other phases than more sophisticated polarizable or _ ab initio _ descriptions .
however partially adiabatic centroid molecular dynamics simulations @xcite have shown that the anharmonic stretch ( eq . [ eq : aharm ] ) allows reasonable agreement to be obtained in the observed frequency shifts in the infrared spectrum in going from liquid to gaseous water as well as from pure light to pure heavy water @xcite .
these models were chosen to assess the importance of anharmonicity in the oh stretch using a `` zeroth order '' description of liquid water .
hence , as we show below , they offer a straightforward way to assess competing quantum effects in water .
figure [ fig : graph ] shows the fractionation factors calculated from our pimd simulations compared to the experimental data of ref .
for consistency with the experimental data , we plot @xmath24 which is simply @xmath25 . since @xmath26 is generally small , @xmath27 thus an experimental value at 280 k of 100 corresponds to 10% more d residing in the liquid than the vapor . above 500 k ,
the experimental data shows a well characterized region of inverse fractionation where d becomes more favored in the vapor than in the liquid , as shown in the inset of fig .
[ fig : graph ] . turning to the simulated data , we observe that the harmonic q - spc / fw model over - predicts the magnitude of fractionation at 300 k by a factor of 3 , and does not fall to the value observed experimentally at 300 k until the temperture is raised to 450 k. in contrast , the q - tip4p / f model is in error by only 25% at the lowest temperature and approaches the experimental values more closely at higher temperatures .
it also correctly shows inverse fractionation above 540 k. a previous study using the rigid spc / e model and @xmath28 perturbation theory , found a h / d fractionation of 450 at 300 k which is @xmath295 times higher than that seen experimentally @xcite .
however , it is not clear whether this was purely due to the use of a rigid model or whether the approximate @xmath28 perturbation technique used to obtain the fractionation ratios was also at fault . while fig.[fig : graph ] demonstrates that the q - tip4p / f model provides much better agreement with the experimental data than q - spcfw , it is not immediately clear what aspect of the parameterization causes this . to better understand the origins of this effect , we constructed two models which we denote aq - spc / fw and hq - tip4p / f .
in the former , the q - spc / fw water model has its harmonic oh stretch replaced by a fourth order morse expansion using the parameters of q - tip4p / f ; in the latter , the morse potential of q - tip4p / f is truncated at the harmonic term ( see eqs .
[ eq : harm ] and [ eq : aharm ] ) . the anharmonic variant of q - spc / fw gives results as good as q - tip4p / f , while the harmonic version of q - tip4p / f fares as poorly as q - spcfw . in other words , the accurate prediction of the fractionation ratios in liquid water
is tied to the anharmonicity in the oh direction and is rather insensitive to the other parameters .
this sheds light on a previous study where a sophisticated rigid polarizable model gave identical predictions for h / d fractionation to the simple fixed - charge rigid spc / e model , i.e. varying the intermolecular potential alone does not give the flexibility required to accurately reproduce the experimental fractionation ratios @xcite . to determine the reason for the inversion observed in fractionation above 500 k in both experiment and the q - tip4p / f model , we decompose the contributions to the fractionation ratio by noting that the quantum contribution to the kinetic energy in eq .
[ eq : kinetic ] is the dot product of two vectors .
the overall kinetic energy is invariant to the coordinate system used to evaluate it , thus when the kinetic energy is calculated in the standard cartesian basis all three components will average to the same number due to the isotropy of the liquid . to gain further insight , we instead use the internal coordinates of the water molecule and determine the contribution to @xmath24 arising from the oh bond vector , a vector in the plane of the molecule , and the vector perpendicular to the molecular plane as shown in fig .
[ fig : waterbeads ] .
this is similar to the approach taken by lin _ et .
_ in the different context of investigating the proton momentum disitribution in ice @xcite .
the results of this decomposition are shown in table [ ta : decompose_300 ] for 300 k where d is experimentally seen to favor the liquid , and table [ ta : decompose_620 ] for 620 k where d is experimentally seen to favor the vapor . from table
[ ta : decompose_300 ] , we see that all of the models have largely similar contributions from the two directions orthogonal to the oh bond and that both are positive and therefore favor the d excess in the liquid .
the contribution perpendicular to the plane is noticeably larger than the contribution in the plane . as shown in eq .
[ eq : deltaa ] , the values depend on the change in the quantum kinetic energy between h and d in the liquid and the vapor , which in turn is determined by how much the h or d atom s position uncertainty is restricted by interacting with the other water molecules in the liquid or vapor . since in the vapor
there is little confinement in plane orthogonal to the water , a larger contribution from that direction is expected ; in the liquid , other molecules are present which restrict expansion in that coordinate . in all cases
the oh contribution is negative , indicating that there is less confinement in the position of the h atom in that direction in the liquid than in the vapor because the hydrogen atom participates in hydrogen bonding allowing the oh chemical bond to stretch more easily .
however , comparing the anharmonic models ( aq - spcfw and q - tip4p / f ) with their harmonic counterparts ( q - spcfw and hq - tip4p / f ) , we observe a 10 fold increase in the values arising from the oh contribution which gives rise to a larger cancellation of the positive contributions from the two orthogonal vectors .
it is this cancellation that leads to the much better agreement with the experimental data at 300 k. we now turn now to table [ ta : decompose_620 ] , which shows the contributions to the fractionation ratio at 620 k , a regime where experimentally d is preferred in the lighter phase .
this is closer to the classical limit where the fractionation would be zero , thus , as expected , each component is reduced in magnitude compared to the lower temperature data in table [ ta : decompose_300 ] .
however , the relative decrease in each component varies .
the contributions arising from the in - plane and out - of - plane contributions orthogonal to oh decrease by a factor of 7 - 8 whereas the oh contribution falls by a factor of only @xmath294 .
for the anharmonic models , the negative oh contribution outweighs the positive components in the other two directions leading to an inversion of the fractionation compared to that seen at 300 k in agreement with the experimental observation of -2 .
the reason the oh component falls off more slowly is that this direction is dominated by stretching of the oh chemical bond , which is a high frequency coordinate , so even at high temperatures , quantum mechanics plays a noticeable role .
in contrast , the two directions orthogonal to the oh direction are lower frequency and so approach the classical limit more rapidly as the temperature is increased .
thus , the reason the h / d fractionation is low around 600 k is not due to the fact that all contributions are individually low but rather that they nearly exactly cancel at this point due to the different rates at which the components approach the classical limit . finally , we computed the fractionation ratio for the ttm3-f water model@xcite , which is known to have a very large cancellation of its quantum mechanical effects at 300 k @xcite .
this model was fit to _ ab initio _ calculations using a potential form incorporating anharmonic flexibility , geometry dependent charges , and polarizability on the oxygen site .
as such , it represents the current gold standard " of parameterized water models and has been used extensively in recent studies probing the effects of quantum mechanical fluctuations on water @xcite . to see
if the large cancellation of quantum effects predicted by this model is consistent with experimental fractionation ratios , we calculated the value at 300 k which yielded a value of @xmath30 .
this is in qualitative and quantitative disagreement with the experimental results , since it predicts d to be favored in the vapor at all temperatures .
this model therefore over - predicts competing quantum effects in water and hence care should be taken concerning its predictions on the effects of quantum fluctuations on water s structure and dynamics .
however , based on our discussion above it is likely that reparamaterization of the oh bond anharmonicity could correct this discrepancy .
in conclusion , we have shown that including anharmonicity in the oh bond when modeling water is essential to obtain agreement with the experimentally observed h / d fractionation ratios and that these ratios provide an excellent method to assess the accuracy of the quantum effects predicted by models of water .
since it has recently been shown that the competition between quantum mechanical effects applies to other hydrogen bonded systems @xcite , it is likely that many of our conclusions will be relevant to understanding isotopic fractionation in these systems .
additionally , while we only considered equilibrium fractionation in this work , which can be calculated exactly for a given potential energy model using pimd simulations , many water processes occurring in the world s atmosphere are non - equilibrium ones . while including the effects of quantum fluctuations on dynamics is a much more challenging feat , the recent development of efficient condensed phase quantum dynamics approaches @xcite should allow insights to be gained into kinetic fractionation processes .
these directions will form the basis of future work .
the authors gratefully thank joseph morrone and david selassie for helpful comments and a critical reading of this manuscript .
this research was supported by a grant to b.j.b . from the national science foundation ( nsf - che-0910943 ) .
vanicek j , miller wh ( 2007 ) efficient estimators for quantum instanton evaluation of the kinetic isotope effects : application to the intramolecular hydrogen transfer in pentadiene .
_ j. chem .
_ 127:114309 .
paesani f , voth ga ( 2010 ) a quantitative assessment of the accuracy of centroid molecular dynamics for the calculation of the infrared spectrum of liquid water . _ the journal of chemical physics _ 132:014105 .
fanourgakis gs , xantheas ss ( 2008 ) development of transferable interaction potentials for water .
v. extension of the flexible , polarizable , thole - type model potential ( ttm3-f , v. 3.0 ) to describe the vibrational spectra of water clusters and liquid water .
_ j. chem .
phys . _ 128:074506 .
liu j , et al .
( 2011 ) insights in quantum dynamical effects in the infrared spectroscopy of liquid water from a semiclassical study with an ab initio - based flexible and polarizable force field .
_ j. chem .
_ 135:244503 . 0.5@cccccc model & stretch & oh & plane & orth & totalq - spcfw & harmonic & -17 & 92 & 141 & 216 hq - tip4p / f & harmonic & -15 & 84 & 146 & 215 q - tip4p / f & anharmonic & -152 & 92 & 164 & 104 aq - spcfw & anharmonic & -149 & 98 & 151 & 100 0.5@cccccc model & stretch & oh & plane & orth & totalq - spcfw & harmonic & -3 & 13 & 20 & 30 hq - tip4p / f & harmonic & -3 & 11 & 20 & 28 q - tip4p / f & anharmonic & -40 & 13 & 22 & -5 aq - spcfw & anharmonic & -40 & 13 & 21 & -6 | when two phases of water are at equilibrium , the ratio of hydrogen isotopes in each is slightly altered due to their different phase affinities .
this isotopic fractionation process can be utilized to analyze water s movement in the world s climate . here
we show that equilibrium fractionation ratios , an entirely quantum mechanical property , also provide a sensitive probe to assess the magnitude of nuclear quantum fluctuations in water . by comparing the predictions of a series of water models ,
we show that those describing the oh chemical bond as rigid or harmonic greatly over - predict the magnitude of isotope fractionation .
models that account for anharmonicity in this coordinate are shown to provide much more accurate results due to their ability to give partial cancellation between inter and intra - molecular quantum effects .
these results give evidence of the existence of competing quantum effects in water and allow us to identify how this cancellation varies across a wide range of temperatures .
in addition , this work demonstrates that simulation can provide accurate predictions and insights into hydrogen fractionation .
ater within earth s atmosphere is naturally composed of the stable hydrogen isotopes hydrogen ( h ) and deuterium ( d ) . during cycles of evaporation ,
condensation and precipitation , these isotopes naturally undergo partial separation due to their differing masses thereby leading to different h / d ratios in the two phases .
this process of fractionation has a number of fortuitous consequences which are utilized in hydrology and geology .
for instance , by comparing the ratio of h to d , one can estimate the origins of a water sample , the temperature at which it was formed , and the altitude at which precipitation occurred @xcite .
equilibrium fractionation , where the two phases are allowed to equilibrate their h / d ratio , is entirely a consequence of the effects of quantum mechanical fluctuations on water s hydrogen bond network .
quantum mechanical effects such as zero - point energy and tunneling are larger for h due to its lower mass . despite numerous studies ,
the extent to which quantum fluctuations affect water s structure and dynamics remains a subject of considerable debate .
it has long been appreciated that one effect of quantum fluctuations in water is the disruption of hydrogen bonding , leading to de - structuring of the liquid and faster dynamics @xcite .
however , more recent work has suggested that a competing quantum effect may exist in water @xcite , namely that the quantum kinetic energy in the oh covalent bond allows it to stretch and form shorter and stronger hydrogen bonds , which partially cancels the disruptive effect .
this hydrogen bond strengthening has only been recently appreciated , as many original studies drew their conclusions based on models with rigid or harmonic bonds which are unable to describe this behavior .
the degree of quantum effect cancellation depends sensitively on the anharmonicity of the oh stretch and the temperature .
these parameters tune the balance between the lower frequency hydrogen bonding disruption , which will dominate at lower temperatures , and the higher frequency hydrogen bond strengthening effect , which will dominate at higher temperatures when rotations become essentially classical .
if such a large degree of cancellation existed at ambient temperature , it would be highly fortuitous both in terms of the biological effects of heavy water , which is only mildy toxic to humans @xcite , as well as the ability to use heavy solvents in 2d - ir and nmr spectroscopies , where deuteration is assumed not to dramatically alter the structure or dynamics observed .
however , the size of this cancellation remains elusive since empirical quantum models of water are typically fit to reproduce its properties when used in path integral simulations and the two _ ab initio _ path integral studies performed have not produced a consistent picture @xcite .
in addition , many of these simulation studies compare the properties of water to those of its classical counterpart , but `` classical '' water is physically unrealizable even at relatively high temperatures , since water still has significant quantum effects present in its vibrations . in this paper , we use equilibrium fractionation ratios as a sensitive probe to assess the magnitude of quantum mechanical effects in water .
fractionation ratios can be directly related to quantum kinetic energy differences between h and d in liquid water and its vapor and can be calculated exactly for a given water potential energy model using path integral simulations .
the large number of accurate experimental measurements of these ratios allows for sensitive comparisons of theory and experiment over a wide range of temperatures @xcite . in the present work ,
we show what features are needed in a water model to accurately predict these ratios by decomposing the contributions to the free energy difference leading to fractionation .
this in turn leads to a simple explanation of the inversion of the fractionation ratios seen experimentally at high temperatures , where d is favored over h in the vapor phase@xcite . |
until 2001 only three candidates existed for @xmath0 hypernuclei observed in emulsion experiments @xcite .
the @xmath0 binding energies deduced from these emulsion events indicated that the @xmath0 interaction is strongly attractive in the @xmath6 channel @xcite , with a @xmath0 pairing energy @xmath7 mev , although it had been realized @xcite that the binding energies of @xmath8be @xcite and @xmath2he @xcite are inconsistent with each other .
this outlook has undergone an important change following the very recent report by the kek hybrid - emulsion experiment e373 of a well - established new candidate @xcite for @xmath2he , with binding energy ( @xmath9 mev ) substantially lower than that deduced from the older , dubious event @xcite .
furthermore , there are also indications from the ags experiment e906 for the production of light @xmath0 hypernuclei @xcite , perhaps as light even as @xmath10h , in the @xmath11 reaction on @xmath12be .
since data on hyperon - nucleon ( @xmath13 ) and hyperon - hyperon ( @xmath14 ) interactions are scarce or even not readily available from laboratory experiments , the study of multistrange systems provides a fairly exclusive test of microscopic models for the baryon - baryon ( @xmath15 ) interaction .
the nijmegen group has constructed over the years a number of one - boson - exchange ( obe ) models ( reviewed by rijken in ref.@xcite and in these proceedings ) for the @xmath15 interaction using su(3)-flavor symmetry to relate coupling constants and phenomenological short - distance hard or soft cores . in all of these rather different @xmath15 interaction models
only 35 @xmath13 low - energy , generally imprecise data points serve the purpose of steering phenomenologically the extrapolation from the @xmath16 sector , which relies on thousands of data points , into the strange @xmath13 and @xmath14 sectors .
it is therefore of utmost importance to confront these models with the new @xmath0 hypernuclear data in order to provide meaningful constraints on the extrapolation to strangeness @xmath17 and beyond .
the @xmath13 and @xmath14 @xmath18-wave interaction input potentials to the structure calculations here reviewed often consist of combinations of gaussians with different ranges , such as to make these single - channel potentials phase equivalent to the nijmegen obe - model coupled - channel potentials .
of the several @xmath0 potentials due to nijmegen models which are shown in fig.[fig : llpot ] , nsc97e is the weakest one , of the order of magnitude required to reproduce @xmath19he ) .
the @xmath0 interaction is fairly weak for all six versions ( @xmath20)-(@xmath21 ) of the nijmegen soft - core model nsc97 @xcite , and versions @xmath22 and @xmath21 provide a reasonable description of single-@xmath23 hypernuclei @xcite .
in this section i will review topical theoretical work on some of the light @xmath0 hypernuclear species connected to old and to new experiments .
the anticipated existence of @xmath2he , now solidly established also experimentally @xcite , leads one to enquire where the onset of @xmath0 binding occurs .
it was argued long ago that the three - body @xmath24 system is unbound @xcite , and hence i will concentrate on the @xmath25 @xmath0 hypernuclear systems . among the few heavier species reported todate ,
@xmath8be will be discussed briefly .
figure [ fig : e6e5 ] demonstrates a nearly linear correlation between faddeev - calculated values of @xmath19he ) and @xmath27h , @xmath26he ) , using several @xmath0 interactions which include ( the lowest - left point ) @xmath28 @xcite . here
@xmath29 where @xmath30 is the @xmath0 binding energy of the hypernucleus @xmath31 and @xmath32 is the ( 2@xmath33 + 1)-average of @xmath34 values for the @xmath35 hypernuclear core levels .
@xmath36 increases monotonically with the strength of @xmath37 , starting in approximately zero as @xmath38 , which is a general feature of three - body models such as the @xmath39 , @xmath40h@xmath0 and @xmath40he@xmath0 models used in these @xmath18-wave faddeev calculations @xcite , and also as shown below for @xmath41 @xmath18-wave faddeev calculations @xcite .
the @xmath42 @xmath43h - @xmath43he hypernuclei are then found to be particle stable for _ all _ the @xmath0 attractive potentials here used .
this conclusion holds also when the @xmath18-wave approximation is relaxed @xcite .
i start by discussing the first faddeev - yakubovsky ( fy ) four - body calculation of @xmath44h @xcite . for two identical hyperons and two essentially identical nucleons ( upon introducing isospin ) as appropriate to a @xmath45 model calculation of @xmath44h , the 18 fy components reduce to seven independent components satisfying coupled equations .
six rearrangement channels are involved in the @xmath18-wave calculation @xcite for @xmath44h(1@xmath46 ) : @xmath47 for 3 + 1 breakup clusters , and @xmath48 with @xmath49=@xmath50+@xmath51 and @xmath52 for the latter 2 + 2 breakup clusters .
using @xmath37 which reproduces @xmath53he ) , the four - body calculation converges well as function of the number @xmath54 of the fy basis functions allowed in , yet it yields no bound state for the @xmath45 system , as demonstrated in fig.[fig : be ] by the location of the ` @xmath45 ' curve _ above _ the horizontal straight line marking the ` @xmath55 threshold . '
@xmath18-wave faddeev calculation which uses model nsc97f @xcite for the underlying @xmath56 interaction , yielding @xmath57 mev . using model nsc97e , with @xmath58 mev ,
does not alter the conclusions listed below .
] in fact these fy calculations exhibit little sensitivity to @xmath37 over a wide range . even for
considerably stronger @xmath0 interactions one gets a bound @xmath44h only if the @xmath56 interaction is made considerably stronger , by as much as 40% . with four @xmath56 pairwise interactions out of a total of six
, the strength of the @xmath56 interaction ( about half of that for @xmath16 ) plays a major role in the four - body @xmath45 problem .
however , fitting a @xmath59 potential to the low - energy parameters of the @xmath18-wave faddeev calculation for @xmath60 and solving the @xmath18-wave faddeev equations for a @xmath61 model of @xmath44h , this latter four - body system is calculated to yield a @xmath62 bound state , as shown in the figure by the location of the asymptote of the ` @xmath61 ' curve _ below _ the ` @xmath55 threshold ' .
the onset of particle stability for @xmath44h(@xmath62 ) requires then a minimum strength for @xmath37 which is exceeded by the choice of @xmath53he ) @xcite as a normalizing datum ( equivalent to @xmath63 fm @xcite ) .
this is demonstrated in fig.[fig : isle ] where faddeev - calculated @xmath64h ) values are shown as function of the @xmath0 scattering length @xmath65 for two different functional forms of the fitted @xmath59 potential . disregarding spin
it can be shown that , for essentially attractive @xmath0 interactions and for a static nuclear core @xmath66 , a two - body @xmath59 bound state implies binding for the three - body @xmath61 system @xcite . however , for a _ non static _ nuclear core @xmath66 ( made out of dynamically interacting proton and neutron ) , a @xmath59 bound state does not necessarily imply binding for the @xmath61 system .
it is questionable whether by incorporating higher partial waves , or @xmath67 coupling effects , this qualitative feature will change .
the above conclusions have been very recently challenged by nemura et al.@xcite .
fig.[fig : akaishi ] demonstrates that within their stochastic variational calculation , which uses the nsc97f input of the filikhin and gal calculation @xcite for the various pairwise interactions , a ` @xmath68 ' model _ always _ yields more binding than a ` @xmath41 ' model does .
particle stability for @xmath44h(@xmath62 ) in this variational calculation requires a minimum strength for @xmath37 which is exceeded by the choice of @xmath53he ) @xcite as a normalizing datum . yet , nemura et al .
argue that the @xmath56 interaction in the @xmath69 channel , when adjusted to the binding energy calculated for the @xmath1 @xmath23 hypernuclei , should be taken weaker than that used by filikhin and gal and that , when this constraint is implemented ( ` set a ' in the figure ) , particle stability for @xmath44h(@xmath62 ) requires a minimum strength for @xmath37 which is not satisfied by the choice of @xmath53he ) as a normalizing datum .
a similar strong dependence on the @xmath69 @xmath56 interaction within a @xmath61 model is shown in fig.[fig : exp ] for two versions @xmath21 and @xmath70 of model nsc97 which produce the same @xmath71 , while differing slightly by the location of the spin - flip @xmath72 excited state @xcite . for heavier @xmath0 hypernuclei ,
the relationship between the three - body and four - body models is opposite to that found by filikhin and gal for @xmath44h : the @xmath74 calculation provides _ higher _ binding than a properly defined @xmath75 calculation yields ( with @xmath76 ) due to the attraction induced by the @xmath77-@xmath78 , @xmath79-@xmath80 , @xmath81-@xmath82 four - body rearrangement channels that include bound states for which there is no room in the three - body @xmath75 model .
the binding energy calculated within the four - body model increases then ` normally ' with the strength of @xmath37 @xcite .
this is demonstrated in fig .
[ fig : e6e10 ] for @xmath73be using several @xmath0 interactions , including @xmath83 which corresponds to the lowest point on each one of the straight lines .
the origin of the dashed axes corresponds to @xmath84 . within the 4-body @xmath85 model , the fairly large value @xmath86 mev in the limit
@xmath38 is due to the special @xmath87 cluster structure of the @xmath88be core .
the correlation noted in the figure between @xmath73be and @xmath89he calculations , and the consistency between various reports on their @xmath90 values , are discussed by filikhin and gal @xcite . in particular , the two solid points next to the lowest one on the ` 4-body model ' line in fig .
[ fig : e6e10 ] , corresponding to two versions of model nsc97 @xcite , are close to reproducing ( the ` new ' ) @xmath53he ) but are short of reproducing ( the ` old ' ) @xmath91be ) by about @xmath92 .
this apparent discrepancy may be substantially reduced by accepting a @xmath73be weak decay scheme that involves the 3 mev excited @xmath93be doublet rather than the @xmath93be ground state @xcite .
this conclusion may also be inferred from the recent 4-body calculations by hiyama et al . for @xmath94 @xmath0 hypernuclei @xcite .
since model nsc97 @xcite provides a qualitatively successful extrapolation from fits to @xmath16 and @xmath13 data to @xmath95 , and noting the strongly attractive @xmath96 @xmath97 potentials shown in fig .
[ fig : pot - ly ] in comparison to the fairly weak @xmath0 potentials when model nsc97 is extrapolated to @xmath98 @xcite , it is natural to search for stability of @xmath99 systems obtained from @xmath89he upon replacing one of the @xmath23 s by @xmath4 .
faddeev calculations @xcite for the @xmath100 @xmath101 ground - state of @xmath102h and @xmath102he , considered as @xmath103 and @xmath104 three - body systems respectively , indicate that @xmath102he is particle - stable against @xmath23 emission to @xmath43he for potentials simulating model nsc97 , particularly versions @xmath22 and @xmath21 , whereas @xmath5h is unstable since @xmath105 by 6.5 mev .
@xmath43h - @xmath43he hypernuclei , within a @xmath75 faddeev calculation , are particle stable even in the limit @xmath38 .
] this is demonstrated in fig .
[ fig : lxi ] . nevertheless ,
predicting particle stability for @xmath102he is not independent of the assumptions made on the experimentally unexplored @xmath106 interaction which was extrapolated from recent data on @xmath107c @xcite ; hence this prediction can not be considered conclusive .
bodmer @xcite , and more specifically witten @xcite , suggested that strange quark matter , with roughly equal composition of @xmath108 , @xmath66 and @xmath18 quarks , might provide the absolutely stable form of matter .
metastable strange quark matter has been studied by chin and kerman @xcite .
jaffe and collaborators @xcite subsequently charted the various scenarios possible for the stability of strange quark matter , from absolute stability down to metastability due to weak decays .
finite strange quark systems , so called strangelets , have also been considered @xcite .
less advertised , perhaps , is the observation made by schaffner et al .
that metastable strange systems with similar properties , i.e. a strangeness fraction @xmath109 and a charge fraction @xmath110 , might also exist in the hadronic basis at moderate values of density , between twice and three times nuclear matter density @xcite .
these strange systems are made out of @xmath54 , @xmath23 and @xmath4 baryons .
the metastability of these strange hadronic systems was established by extending relativistic mean field ( rmf ) calculations from ordinary nuclei ( @xmath111 ) to multi - strange nuclei with @xmath112 .
although the detailed pattern of metastability , as well as the actual values of the binding energy , depend specifically on the partly unknown hyperon potentials assumed in dense matter , the predicted phenomenon of metastability turned out to be robust in these calculations @xcite .
recently , model nsc97 and its extension @xcite were used to calculate within the rmf framework the minimum - energy equilibrium composition of bulk strange hadronic matter ( shm ) made out of the su(3 ) octet baryons @xmath113 and @xmath4 , over the entire range of strangeness fraction @xmath114 @xcite .
the main result is that shm is comfortably metastable in this model for any allowed value of @xmath115 .
the @xmath116 composition and the binding energy calculated for equilibrium configurations with @xmath117 resemble those of model 2 in refs .
the extension of model nsc97 @xcite yields particularly attractive @xmath118 , @xmath119 and @xmath120 interactions , but fairly weak @xmath121 and @xmath122 interactions .
consequently , for @xmath123 , @xmath124 s replace @xmath23 s due to their exceptionally strong attraction to @xmath124 and @xmath4 hyperons .
as is shown below , a first - order phase transition occurs from @xmath125 dominated matter for @xmath117 to @xmath126 dominated matter for @xmath123 , with binding energies per baryon reaching as much as 80 mev . a phase transition is visualized in fig .
[ fig : sbg4 ] where the binding energy is drawn versus the baryon density for several representative fixed values of @xmath127 . for @xmath128 , there is a global minimum at a baryon density of @xmath129 @xmath130 . a shallow local minimum
is seen at larger baryon density at @xmath131 @xmath130 .
increasing the strangeness fraction to @xmath132 lowers substantially the local minimum by about 20 mev , whereas the global minimum barely changes . at @xmath133 this trend
is amplified and the relationship between the two minima is reversed , as the minimum at higher baryon density becomes energetically favored .
the system will then undergo a transition from the low density state to the high density state .
due to the barrier between the two minima , it is a first - order phase transition from one minimum to the other .
[ fig : sbg5 ] demonstrates explicitly that the phase transition involves transformation from @xmath134 dominated matter to @xmath135 dominated matter , by showing the calculated composition of shm for this model ( denoted n ) as function of the strangeness fraction @xmath127 .
the particle fractions for each baryon species change as function of @xmath127 . at @xmath136 ,
one has pure nuclear matter , whereas at @xmath137 one has pure @xmath4 matter . in between , matter is composed of baryons as dictated by chemical equilibrium .
a change in the particle fraction may occur quite drastically when new particles appear , or existing ones disappear in the medium .
a sudden change in the composition is seen in fig .
[ fig : sbg5 ] for @xmath138 when @xmath4 s emerge in the medium , or at @xmath139 when nucleons disappear .
the situation at @xmath140 is a special one , as @xmath124 s appear in the medium , marking the first - order phase transition observed in the previous figure .
the baryon composition alters completely at that point , from @xmath141 baryons plus a rapidly vanishing fraction of @xmath23 s into @xmath142 hyperons plus a decreasing fraction of nucleons . at the very deep minimum of the binding energy curve ( fig .
@xcite ) shm is composed mainly of @xmath124 s and @xmath4 s with a very small admixture of nucleons .
i have presented faddeev calculations for @xmath43h - @xmath43he and @xmath89he , and first ever four - body faddeev - yakubovsky calculations for @xmath44h and @xmath73be , using two - body potentials fitted to the low - energy scattering parameters or to the binding energies of the respective subsystems . in particular , for @xmath44h , @xmath16 and @xmath56 interaction potentials that fit the binding energy of @xmath143h were used .
no @xmath44h bound state was obtained for a wide range of @xmath0 interactions , including that corresponding to @xmath53he ) .
this non binding is due to the relatively weak @xmath56 interaction , in stark contrast to the results of a ` reasonable ' three - body @xmath61 faddeev calculation .
further experimental work is needed to decide whether or not the events reported in the ags experiment e906 @xcite correspond to @xmath44h , particularly in view of subsequent conflicting theoretical analyses @xcite .
more theoretical work , particularly on the effects of including explicitly @xmath144 channel couplings , is called for .
preliminary estimates for such effects within the nsc97 model , or its simulation , have been recently made @xcite .
in particular , lanskoy and yamamoto @xcite have focused attention to the substantial charge symmetry breaking effects introduced by the different @xmath145he thresholds into the binding energy calculation of the @xmath43h - @xmath43he ground states .
in addition to increasing the calculated @xmath36 value of @xmath43he with respect to that of @xmath43h , on top of the difference already shown in fig . [
fig : e6e5 ] , it is found that for model nsc97 @xmath146he ) gets as large and perhaps even larger than that of @xmath89he ( @xmath147 mev ) .
a simultaneous determination of the binding energies of @xmath43he and @xmath89he would help to discriminate between several versions of obe models which differ markedly from each other regarding the strength of the off - diagonal @xmath67 coupling . accepting the predictive power of model nsc97
, faddeev calculations suggest that @xmath5he may be the lightest particle - stable @xmath148 hypernucleus , and the lightest and least strange particle - stable hypernucleus in which a @xmath4 hyperon is bound .
unfortunately , the direct production of @xmath97 hypernuclei is beyond present experimental capabilities , requiring the use of @xmath149 initiated reactions .
finally , i have focused on the consequences of using model nsc97 for the binding and composition of strange hadronic matter .
strange hadronic matter is comfortably stable , up to weak decays , over a wide range of baryon - baryon interaction models , including model nsc97 here chosen because it successfully extrapolates from the @xmath150 sectors in which it was constructed into the @xmath95 sector , nearly reproducing @xmath53he ) .
the phase transition considered in this review has been recently discussed by the frankfurt group @xcite in the context of phase transition to hyperon matter in neutron stars .
unfortunately , it will take lots of imagination to devise experimentally a way to determine how attractive those @xmath97 , @xmath151 , @xmath152 , @xmath153 interactions are , which are so crucial for the results exhibited in this review . | recent experiments on production of @xmath0 hypernuclei have stimulated renewed interest in extracting the @xmath0 interaction from the few events identified since the inception of this field forty years ago .
few - body calculations relating to this issue are reviewed , particularly with respect to the possibility that @xmath1 marks the onset of @xmath0 binding to nuclei .
the nijmegen soft - core model potentials nsc97 qualitatively agree with the strength of the @xmath0 interaction deduced from the newly determined binding energy of @xmath2he . applying the extended nsc97 model to stranger nuclear systems suggests that @xmath3 marks the onset of @xmath4 binding , with a particle stable @xmath5he , and that strange hadronic matter is robustly bound . |
theoretical investigations of systems with strong electron correlations in the magnetic field is of interest in connection with the observation of quantum oscillations in the mixed state of underdoped cuprates @xcite .
based on the onsager - lifshitz - kosevich theory for metals @xcite the measured decreased quantum oscillation frequencies were interpreted as a manifestation of small fermi surface pockets @xcite . if one anticipate that the normal state in the field coincides with the zero - field state , this interpretation seems to contradict numerous photoemission experiments . to explain the appearance of these small pockets proposals for various states with broken translational symmetry
were suggested @xcite .
other theories relate the decreased quantum oscillation frequency to superconducting fluctuations @xcite or use phenomenology of the marginal fermi liquid @xcite .
crystals , in which the decreased quantum oscillation frequencies were observed , belong to the underdoped region of the cuprate phase diagram , and , therefore , they are characterized by strong electron correlations . the behavior of models of strong correlations in magnetic fields is poorly known . in this work we consider the two - dimensional ( 2d ) @xmath0-@xmath1 model of cu - o planes .
the used approach allows us to take proper account of strong electron correlations and to consider large enough clusters in moderate magnetic fields .
an approximation is exploited , in which normal - state holes are characterized by large zero - field fermi surfaces for hole concentrations @xmath5 .
we found that in the range @xmath3 the density of hole states at the fermi level oscillates with the inverse magnetic field @xmath2 .
the oscillations have components with frequencies differing by an order of magnitude .
the high - frequency components are connected with the large fermi surfaces , while the low - frequency components are related to van hove singularities in the landau subbands , which traverse the fermi level with changing @xmath4 .
these van hove singularities are linked with bending the landau subbands due to strong correlations .
frequencies of slow components are of the same order of magnitude as quantum oscillation frequencies observed in underdoped cuprates .
hence , the @xmath0-@xmath1 model supplemented with a mechanism , which smears out high - frequency oscillations , is able to interpret low frequencies of quantum oscillations observed in the experiments .
the hamiltonian of the 2d @xmath0-@xmath1 model in the perpendicular magnetic field reads @xmath6 where 2d vectors @xmath7 and @xmath8 label sites of a square plane lattice , @xmath9 is the projection of the hole spin , @xmath10 and @xmath11 are hole creation and annihilation operators with the empty @xmath12 and singly occupied @xmath13 site states .
these three states form the complete set of hole states for the site * l * in the @xmath0-@xmath1 model .
the hole kinetic energy @xmath14 contains the hopping matrix element @xmath15 and the peierls phase factor @xcite with the vector potential @xmath16 .
the exchange term includes the exchange constant @xmath17 and the spin-@xmath18 operators @xmath19 and @xmath20 of localized spins .
the zeeman term is omitted , since for the considered fields and exchange constants in cuprates it is two orders of magnitude smaller than the exchange term . in the following consideration
we suppose that only nearest neighbor hopping and exchange constants are nonzero , @xmath21 , @xmath22 where @xmath23 are four vectors connecting nearest neighbor sites . in the landau gauge @xmath24 , where @xmath25 is the @xmath26 component of the site vector @xmath7 and @xmath27 is the unit vector along the @xmath28 axis . hence the exponential in the kinetic term of the hamiltonian can be written as @xmath29 we shall restrict our consideration to the fields satisfying the condition @xmath30 where @xmath31 , @xmath32 and @xmath33 are integers with no common factor . in this case
the kinetic term of the hamiltonian defines its translation properties
@xmath14 is invariant with respect to translations by the lattice period along the @xmath28 axis and by @xmath32 lattice periods along the @xmath26 axis . to retain this symmetry we apply the periodic born - von karman boundary conditions to the sample with @xmath34 sites along the @xmath28 axis and @xmath35 sites along the @xmath26 axis .
the boundary conditions define the set of allowed wave vectors with components @xmath36 and @xmath37 with integer @xmath38 and @xmath39 .
as can be seen from ( [ pphase ] ) and ( [ condition ] ) , the momenta @xmath40 coincide with some wave vectors in this net .
this allows us to perform the usual fourier transformation , @xmath41 , @xmath42 , and obtain for @xmath14 @xmath43 it is convenient to split the brillouin zone into @xmath32 stripes of the width @xmath44 , which are oriented parallel to the @xmath28 axis .
if we select the lowest stripe with @xmath45 , and denote wave vectors in it as @xmath46 , momenta in the entire brillouin zone can be described as @xmath47 . here
@xmath48 and @xmath49 . in these notations
the kinetic energy acquires the form @xmath50 , where the summation over @xmath46 is performed over the selected stripe , @xmath51 and the matrix @xmath52 has the following elements : @xmath53 in this equation the matrix indices @xmath54 and @xmath55 are determined modulo @xmath32 .
the hermitian matrix ( [ hmatr ] ) can be diagonalized by the unitary transformation @xmath56 .
since the kinetic energy defines symmetry properties of the total hamiltonian ( [ hamiltonian ] ) , the operators @xmath57 form a basis for a representation of the symmetry group of the hamiltonian .
besides , in the absence of correlations states created by @xmath58 are eigenstates of the hamiltonian . therefore , green s function constructed from these operators , @xmath59 is an appropriate mean for investigating the influence of strong correlations on hole states . in ( [ green ] ) , the averaging over the grand canonical ensemble and the operator time dependence are determined by the hamiltonian @xmath60 with the chemical potential @xmath61 . to calculate ( [ green ] ) we use the mori projection operator technique @xcite . in this approach ,
the fourier transform of green s function ( [ green ] ) is represented by the continued fraction @xmath62 where @xmath63\nonumber\\ & & \qquad -\frac{3jf_1}{\phi}\sum_{j=0}^{n-1}|u_{{\bf k}jm}|^2\gamma_{{\bf k}+j{\boldsymbol{\kappa}}}-\frac{3jc_1}{\phi}-\mu,\label{e0}\\ & & v_0=\sum_{j=0}^{n-1}\sum_{\bf aa'}u^*_{{\bf k}jm}u_{{\bf k},j+j_{\bf a}-j_{\bf a'},m}{\rm e}^{i({\bf k}+j{\boldsymbol{\kappa}})({\bf a - a'})-i{\boldsymbol{\kappa}}_{\bf a}{\bf a'}}\nonumber\\ & & \qquad\times t^2\bigg(\phi^2+\frac{3c_1}{2\phi}+\frac{3c_{\bf a - a'}}{2}\bigg)\nonumber\\ & & \qquad -t(\mu+e_0)\sum_{j=0}^{n-1}\sum_{\bf a}u^*_{{\bf k}jm}u_{{\bf k},j+j_{\bf a},m}\bigg[{\rm e}^{i({\bf k}+j{\boldsymbol{\kappa}}){\bf a } } \bigg(2\phi+\frac{3c_1}{\phi}\bigg)+\frac{2f_1}{\phi}\bigg]\nonumber\\ & & \qquad+\frac{t^2f_1}{\phi}{\rm re}\sum_{j=0}^{n-1}\sum_{\bf aa'}u^*_{{\bf k}jm}u_{{\bf k},j+j_{\bf a}-j_{\bf a'},m}{\rm e}^{i({\bf k}+j{\boldsymbol{\kappa}}){\bf a}}\nonumber\\ & & \qquad -\frac{t^2f_1}{\phi}\sum_{j=0}^{n-1}\sum_{\bf
a}u^*_{{\bf k}jm}u_{{\bf k},j+2j_{\bf a},m}{\rm e}^{i({\bf k}+j{\boldsymbol{\kappa}}){\bf a } } \nonumber\\ & & \qquad + ( \mu+e_0)^2+t^2x-\frac{4t^2c_1}{\phi},\label{v0}\\ & & e_1\approx -\mu .
\label{e1}\end{aligned}\ ] ] in the above equations , @xmath64 , @xmath65 , @xmath66 is the hole concentration , @xmath67 , @xmath68 @xmath69 and @xmath70 . in the course of calculations parameters @xmath28 and @xmath71 were determined self - consistently .
parameters @xmath72 and @xmath73 , which characterize the spin subsystem and are indirectly influenced by the field , were taken from the results of zero - field calculations @xcite .
the quantity @xmath74 is described by a more complex expression than that given by ( [ e1 ] ) .
the comparison of results obtained with this expression and with ( [ e1 ] ) shows their similarity .
therefore , to speed up calculations we used the latter value for @xmath74 . notice that ( [ hmatr ] ) , ( [ e0 ] ) and ( [ v0 ] ) generalize equations of ref .
@xcite to the case @xmath75 , which allows us to vary @xmath2 with a step as small as is wished .
terminating calculations of the continued fraction ( [ cfraction ] ) with the term @xmath74 green s function ( [ green ] ) is approximated by the expression with two poles for every wave vector @xmath46 and subband index @xmath76 .
this procedure gives the first mori correction to the uncorrelated solution described by the operator @xmath58 .
the correction takes into account electron correlations . in view of rapid convergence of utmost poles of a finite continued fraction with the number of its terms ,
this procedure is supposed to give a qualitatively correct description for these poles .
this supposition is supported by the results obtained with this approximation for @xmath77 @xcite . in this case
the pole with the lower energy corresponds to the spin - polaron band , which gives the most intensive maximum in the hole spectral function .
the two - pole approximation provides a satisfactory description of the evolution of this band with doping . for @xmath78
the states near the fermi level originate from this band , which gives grounds to use the same approximation for their description .
notice , however , that the comparatively small number of calculated continued fraction elements does not allow us to describe the pseudogap and fermi arcs in the zero - field hole spectrum . in the used approximation
the @xmath77 fermi surface is large for @xmath79 , having shape of diamonds with rounded corners , which are centered at @xmath80 . in the @xmath0-@xmath1 model , at moderate doping the pseudogap and fermi arcs reflect the dispersions of the spin - polaron band with the bandwidth @xmath81 and the spectrally less intensive band with the bandwidth @xmath82 @xcite .
this latter band has the mentioned large fermi surface .
therefore , we suppose that the used approximation gives at least qualitatively a correct picture of the energy spectrum at @xmath78 .
for @xmath78 the density of the hole states ( dos ) , @xmath83 as a function of frequency and magnetic field reveals oscillations near the fermi level .
an example of these oscillations is shown in fig .
[ fig1 ] . ,
@xmath84 and @xmath85 .
the fermi level is shown by the red dashed line . ] here and hereafter we set @xmath0 and @xmath86 as units of energy and length , respectively . in the calculations we set @xmath87 and the temperature @xmath88 .
the oscillations are observed in the concentration range @xmath3 . outside of this range the fermi level falls on strong maxima of the dos , where the oscillations are lost against the background . for @xmath89 and @xmath85 .
calculated values are shown by symbols , connecting lines are a guide to the eye .
( b ) the same data smoothed by the fft filter method . ] as mentioned above , in the considered range of @xmath28 the zero - field fermi surfaces are large . the calculated period of oscillations of @xmath90 as a function of @xmath91 conforms with such fermi surfaces for the case @xmath85 the period is equal to @xmath92 [ see fig .
[ fig2](a ) ] in the units of @xmath93 , which corresponds to the onsager frequency @xmath94 kt for @xmath95 (an approximate distance between copper sites in cu - o planes ) . however , along with these high - frequency oscillations a modulation with a period which is larger by an order of magnitude is also observed . to reveal the respective low - frequency oscillations
the data were smoothed with the fft filter method , which allows one to suppress high - frequency components .
the result is shown in fig .
[ fig2](b ) .
the frequency of these oscillations is of the order of 1 kt , which is comparable to the dominant frequency of quantum oscillations in underdoped yba@xmath96cu@xmath97o@xmath98 @xcite .
thus , supplemented with a mechanism , which smears out high - frequency oscillations , the @xmath0-@xmath1 model is able to explain the low - frequency quantum oscillations observed in cuprate perovskites . as candidates for such a mechanism finite lifetimes of hole states , hole spins , field and crystal inhomogeneities @xcite can be mentioned . for other considered hole concentrations in the range @xmath3 similar pictures of high- and low - frequency oscillations in the dos can be revealed .
( a ) , 52 ( b ) and 56 ( c ) . @xmath85 and @xmath84 .
the fermi level is shown by the purple ( dark ) contour.,title="fig : " ] ( a ) , 52 ( b ) and 56 ( c ) . @xmath85 and @xmath84 .
the fermi level is shown by the purple ( dark ) contour.,title="fig : " ] ( a ) , 52 ( b ) and 56 ( c ) . @xmath85 and @xmath84 . the fermi level is shown by the purple ( dark ) contour.,title="fig : " ] what is the reason for the appearance of the low - frequency modulation in fig .
[ fig2](a ) and oscillations in fig .
[ fig2](b ) ? in the uncorrelated case carrier energies form a stair of nearly * k*-independent landau subbands ( in which energies experience weak oscillations when @xmath99 varies from @xmath100 to @xmath101 at a fixed @xmath102 in the landau gauge ) .
the traverse of the fermi level through these subbands leads to oscillations in the dos with a nearly constant amplitude .
strong correlations lead to a bend of the subbands along the @xmath28 direction .
this bending is seen in fig .
[ fig3 ] . in this figure ,
only the subbands near the fermi level are shown for the case @xmath84 .
the subband @xmath76 is plotted in the stripe @xmath103 , @xmath104 of the entire brillouin zone . as a result
the hole dispersion looks like stairs with steps in the @xmath26 direction .
these stairs ascend from smaller hole energies to larger ones with @xmath105 moving from @xmath100 to @xmath101 .
the small - scale oscillations along the @xmath28 direction have the same origin as in the uncorrelated case .
bends of the subbands lead to the appearance of van hove singularities , and these singularities supplement the dos oscillations with the amplitude modulation seen in fig . [ fig2](a ) .
panels of fig .
[ fig3 ] demonstrate cases when the fermi level falls onto a van hove singularity [ panel ( c ) ] , is halfway between two singularities [ panel ( a ) ] and is located in some intermediate position [ panel ( b ) ] .
thus , the low - frequency modulation and oscillations in fig .
[ fig2 ] are connected with the traverse of the fermi level through the sequence of van hove singularities in the landau subbands .
in summary , we have considered the two - dimensional @xmath0-@xmath1 model of cu - o planes of cuprate perovskites under the conditions of strong electron correlations : @xmath106 and small hole concentrations @xmath28 .
the two - dimensional crystal is placed in a magnetic field , which is perpendicular to the crystal plane . using the mori projection operator technique
we have calculated the landau subbands in the case when only the nearest neighbor hopping constant is nonzero and for the magnetic field induction satisfying the condition @xmath107 , where @xmath86 is the lattice spacing , @xmath32 and @xmath33 are integers with no common factor . in the range of hole concentrations @xmath3 the density of hole states at the fermi level @xmath90 shows oscillations as a function of @xmath2 . for somewhat smaller and larger hole concentrations
the fermi level falls onto strong maxima of the density of states , which hide the oscillations .
the oscillations have high- and low - frequency components .
frequencies of the former components conform with large zero - field fermi surfaces , which are inherent in the model for these @xmath28 in the used approximation .
the components with an order of magnitude smaller frequencies are connected with the bending of the landau subbands near the fermi level .
this bending is a result of strong electron correlations .
the bending leads to the appearance of van hove singularities in the landau subbands , which traverse the fermi level with changing @xmath4 .
onsager frequencies of these components @xmath108 kt are of the same order of magnitude as dominant quantum oscillation frequencies observed in underdoped cuprate perovskites . being supplemented with a mechanism , which smears out the high - frequency components , the calculated oscillations of @xmath90 become similar to quantum oscillations in these experiments .
this work was supported by the european regional development fund ( project tk114 ) and by the estonian scientific foundation ( grant etf9371 ) .
99 n. doiron - leyraud , c. proust , d. leboeuf , j. levallois , j .- b .
bonnemaison , ruixing liang , d.a .
bonn , w.n .
hardy , l. taillefer , nature * 447 * , 565 ( 2007 ) .
bangura , p.a .
goddard , j. singleton , s.w .
tozer , a.i .
coldea , a. ardavan , r.d .
mcdonald , s.j .
blundell , j.a .
schlueter , phys .
b * 76 * , 052510 ( 2007 ) .
yelland , j. singleton , c.h .
mielke , n. harrison , f.f .
balakirev , b. dabrowski , j.r .
cooper , phys .
lett . * 100 * , 047003 ( 2008 ) .
d. shoenberg , _ magnetic oscillations in metals _
( cambridge university press , cambridge , 1984 ) .
sebastian , n. harrison , g.g .
lonzarich , rep .
. phys . * 75 * , 102501 ( 2012 ) .
a.j . millis , m. norman , m. , phys . rev .
b * 76 * , 220503(r ) ( 2007 ) . w .-
chen , k .- y .
yang , t.m .
rice , f.c .
zhang , epl * 82 * , 17004 ( 2008 ) .
v. galitski , s. sachdev , phys .
b * 79 * , 134512 ( 2009 ) .
a. melikyan , o. vafek , phys .
b * 78 * , 020502(r ) ( 2008 )
. t. pereg - barnea , h. weber , g. rafael , m. franz , nature phys . * 6 * , 44 ( 2010 )
varma , phys .
b * 79 * , 085110 ( 2009 ) .
r. peierls , z. phys .
* 80 * , 763 ( 1933 ) .
h. mori , progr .
* 34 * , 399 ( 1965 ) .
a. sherman , phys .
b * 70 * , 184512 ( 2004 ) . a. sherman , m. schreiber , phys . lett .
a * 377 * , 2979 ( 2013 ) .
a. sherman , m. schreiber , phys .
b * 55 * , r712 ( 1997 ) . | the normal - state energy spectrum of the two - dimensional @xmath0-@xmath1 model in a homogeneous perpendicular magnetic field is investigated .
the density of states at the fermi level as a function of the inverse magnetic field @xmath2 reveals oscillations in the range of hole concentrations @xmath3 .
the oscillations have both high- and low - frequency components .
the former components are connected with large fermi surfaces , while the latter with van hove singularities in the landau subbands , which traverse the fermi level with changing @xmath4 .
the singularities are related to bending the landau subbands due to strong electron correlations .
frequencies of these components are of the same order of magnitude as quantum oscillation frequencies observed in underdoped cuprates . |
first - passage percolation theory on periodic graphs was presented by hammersley and welsh @xcite to model the spread of a fluid through a porous medium . in this paper
we continue a study of planar first - passage percolation models on random graphs , initiated by vahidi - asl and wierman @xcite , as follows .
let @xmath12 denote the set of points realized in a two - dimensional homogeneous poisson point process with intensity @xmath13 .
to each @xmath14 corresponds an open polygonal region @xmath15 , the voronoi tile at @xmath16 , consisting of the set of points of @xmath17 which are closer to @xmath16 than to any other @xmath18 .
given @xmath19 we denote by @xmath20 the almost surely unique point in @xmath12 such that @xmath21 .
the collection @xmath22 is called the voronoi tiling of the plane based on @xmath12 .
the delaunay triangulation @xmath0 is the graph where the vertex set @xmath23 equals @xmath12 and the edge set @xmath24 consists of non - oriented pairs @xmath25 such that @xmath26 and @xmath27 share a one - dimensional edge ( figure [ f : vor2 ] ) .
one can see that almost surely each voronoi tile is a convex and bounded polygon , and the graph @xmath0 is a triangulation of the plane @xcite .
the voronoi tessellation @xmath28 is the graph where the vertex set @xmath29 is the set of vertices of the voronoi tiles and the edge set @xmath30 is the set of edges of the voronoi tiles .
the edges of @xmath28 are segments of the perpendicular bisectors of the edges of @xmath0 .
this establishes duality of @xmath0 and @xmath28 as planar graphs : @xmath31 .
to each edge @xmath32 is independently assigned a nonnegative random variable @xmath33 from a common distribution @xmath3 , which is also independent of the poisson point process that generates @xmath12 . from now on we denote @xmath34 the probability space induced by the poisson point process @xmath12 and the passage times @xmath35 . the passage time @xmath36 of a path @xmath37 in the delaunay triangulation is the sum of the passage times of the edges in @xmath37 .
the first - passage time between two vertices @xmath16 and @xmath38 is defined by @xmath39 where @xmath40 the set of all paths connecting @xmath16 to @xmath38 . given @xmath41 we define @xmath42 . to state the main result of this work
we require some definitions involving a bond percolation model on the voronoi tessellation @xmath28 .
such a model is constructed by choosing each edge of @xmath28 to be open independently with probability @xmath43 .
an open path is a path composed of open edges .
we denote @xmath44 the law induced by the poisson point process and the random state ( open or not ) of an edge . given a planar graph @xmath45 and @xmath46 we say that a self - avoiding path @xmath47 is a path connecting @xmath48 to @xmath49 if @xmath50\cap\ba\neq\emptyset$ ] and @xmath51\cap\bb\neq\emptyset$ ] ( @xmath52 $ ] denotes the line segment connecting @xmath4 to @xmath53 ) .
for @xmath54 let @xmath55 be the event that there exists an open path @xmath56 in @xmath28 , connecting @xmath57 $ ] to @xmath58 $ ] , and with @xmath59\times [ 0,l]$ ] for all @xmath60 . in this case
we also say that @xmath37 crosses the rectangle @xmath61\times [ 0,l]$ ] .
define the function @xmath62 and consider the percolation threshold , @xmath63 we have that @xmath64 , which follows by standard arguments in percolation theory . for
more in percolation thresholds on voronoi tilings we refer to @xcite .
[ t1 ] if @xmath65 then there exist constants @xmath66 such that for all @xmath67 @xmath68 where @xmath69 and @xmath70 . to show the importance of theorem [ t1 ] we recall two fundamental results proved by vahidi - asl and wierman @xcite . consider the growth process @xmath71 where @xmath72 denotes the closure of @xmath73 .
set @xmath74\,.\ ] ] and let @xmath75 be independent random variables with distribution @xmath3 . if @xmath76 then @xmath77 and for all unit vectors @xmath78 ( @xmath79 ) @xmath80-a.s .
@xmath81 further , if @xmath82 and @xmath9 then for all @xmath83 @xmath80-a.s . there exists @xmath84 such that for all @xmath85 @xmath86 where @xmath87 .
we note here that the asymptotic shape is an euclidean ball due to the statistical invariance of the poisson point process .
unfortunately the exact value of the time constant @xmath7 , as a functional of @xmath3 , still a basic problem in first - passage percolation theory .
our result provides a sufficient condition on @xmath3 to ensure @xmath9 .
[ col1 ] under assumption , if @xmath65 then @xmath88 . together with the borel - cantelli lemma , theorem [ t1 ] and
imply @xmath89 which is the desired result .
for fpp models on the @xmath90 lattice kesten ( 1986 ) have shown that @xmath91 ( the critical probability for bond percolation on @xmath90 ) is a sufficient condition to get by using a stronger version of the bk - inequality . here
we follow a different method and apply a simple renormalization argument to obtain a similar result .
we expect that our condition to get is equivalent to @xmath92 where @xmath93 is the probability that bond percolation on @xmath0 occurs with density @xmath43 , since it is conjectured that @xmath94 ( duality ) for many planar graphs .
in fact , by combining corollary [ col1 ] with we have : [ e1cr ] @xmath95 to see this assume we have a first - passage percolation model on @xmath0 with @xmath96 then @xmath80-a.s .
there exists an infinite cluster @xmath97 composed by edges @xmath1 with @xmath98 .
denote by @xmath99 the first - passage time from @xmath5 to @xmath100 .
then for all @xmath101 we have that @xmath102 is an unbounded set . by ( since such a distribution satisfies and ) , this implies that @xmath103 if @xmath104 . on the other hand , by corollary [ col1 ] , @xmath105 if @xmath106 , and so must hold .
other passage times have been considered in the literature such as @xmath107 , where @xmath108 is the hyperplane consisting of points @xmath109 so that @xmath110 , and @xmath111 ^ 2)$ ]
. the arguments in this article can be used to prove the analog of theorem [ t1 ] when @xmath112 is replaced by @xmath107 or @xmath111 ^ 2)$ ] . for site versions of fpp models
the method works as well if we change the condition on @xmath3 to @xmath113 , where now @xmath114 is the critical probability for site percolation . similarly
to corollary [ e1cr ] , in this case one can also obtain the inequality @xmath115 . for more details
we refer to @xcite .
for the moment we assume that @xmath3 is bernoulli with parameter @xmath43 .
let @xmath116 be a parameter whose value will be specified later .
let @xmath117 and @xmath118 denote @xmath119 the circuit composed by sites @xmath120 with @xmath121 .
for each @xmath122 , we denote by @xmath123 its boundary . for each @xmath124 and
@xmath125 consider the box @xmath126^{2}\,.\ ] ] divide @xmath127 into thirty - six sub - boxes with the same size and declare that @xmath128 is a * full box * if all these thirty - six sub - boxes contain at least one point of @xmath12 .
let @xmath129\,.\ ] ] let @xmath130 be the set of all self - avoiding paths @xmath56 in @xmath0 , connecting @xmath131 to @xmath132 and with @xmath133 for all @xmath60 .
let @xmath134\,.\ ] ] we say that @xmath127 is a * good box * ( or that @xmath135 is a good point ) if @xmath136 where @xmath137 denotes the indicator function of the event @xmath138 . now , let @xmath143 , where @xmath144 is the edge in @xmath30 ( the voronoi tessellation ) dual to @xmath1 . then @xmath145 defines a bond percolation model on @xmath28 with law @xmath44 .
consider the rectangles @xmath146\times[-3l/2,3l/2]\,,\,r^2_l:=[-3l/2,3l/2]\times[l/2,3l/2]\ ] ] @xmath147\times[-3l/2,3l/2]\mbox { and } r_l^4:=[-3l/2,3l/2]\times[-3l/2,-l/2]\,.\ ] ] we denote by @xmath148 the event @xmath149 ( recall the definition of @xmath10 ) but now translate to the rectangle @xmath150 , and by @xmath151 the event that an open circuit @xmath152 in @xmath28 which surrounds @xmath153 and lies inside @xmath154 does not exist .
thus one can easily see that @xmath155 notice that if there exists an open circuit @xmath152 in @xmath28 which surrounds @xmath153 and lies inside @xmath154 , then every path @xmath37 in @xmath130 has an edge crossing with @xmath152 and thus @xmath156 .
therefore , @xmath157 since @xmath158 , by using , , and the definition of @xmath159 , we get lemma [ l0 ] . to obtain some sort of independence between the random variables
@xmath160 we shall study some geometrical aspects of voronoi tilings .
given @xmath161 , let @xmath162 be the sub - graph of @xmath0 composed of vertices @xmath163 in @xmath23 and edges @xmath164 in @xmath24 so that @xmath165 for all @xmath166 .
[ l2 ] let @xmath54 and @xmath167 .
assume that @xmath12 and @xmath168 are two configurations of points so that @xmath169 and that @xmath170 is a full box with respect to @xmath12 , for all @xmath171
. then @xmath172 . by the definition of the delaunay triangulation , lemma [ l2 ] holds if we prove that @xmath173 to prove this we claim that @xmath174 if does not hold then there exist @xmath175 and @xmath176 ( by convexity of voronoi tilings ) .
since every box @xmath177 with @xmath178 is a full box , there exist @xmath179 so that @xmath180 although , @xmath181 and @xmath182 belong to @xmath183 and so @xmath184 thus , @xmath185 which leads to a contradiction since @xmath186 . by an analogous argument
, one can prove that @xmath187 now suppose does not hold . without lost of generality
, we may assume that there exists @xmath14 with @xmath188 and @xmath189 with @xmath190 .
so @xmath191 for some @xmath192 .
although , @xmath169 and then @xmath193 , which is a contradiction with and .
for each @xmath194 , we say that the collection of random variables @xmath195 is @xmath196-dependent if @xmath197 and @xmath198 are independent whenever @xmath199 combining lemma [ l2 ] with the translation invariance and the independence property of the poisson point process we obtain : combining lemmas [ l0 ] and [ l1 ] with and theorem 0.0 of ligget , schonman and stacey @xcite , one gets that @xmath209 is dominated from below by a collection @xmath210 of i.i.d .
bernoulli random variables with parameter @xmath211 when @xmath212 .
but for @xmath213 sufficiently close to @xmath13 , we can chose @xmath214 sufficiently small , so that the probability of the event that @xmath215 decays exponentially fast with @xmath216 ( see chapter 3 of grimmett @xcite ) . together with domination
, this proves lemma [ l3 ] .
we say that @xmath219 is a circuit of good boxes if @xmath220 is a good circuit in @xmath204 , and that @xmath219 and @xmath221 are @xmath196-distant if @xmath222 denote @xmath223 .
notice that there exist at least @xmath224 pairwise @xmath201-distant circuits of good boxes surrounding the origin and lying inside @xmath225 ^ 2\subseteq\r^2 $ ] .
therefore , every path @xmath37 between the origin and any point outside @xmath225 ^ 2 $ ] must cross at least @xmath224 @xmath201-distant circuits of good boxes .
we claim this yields @xmath226 indeed , assume we take two @xmath201-distant good boxes , say @xmath227 and @xmath228 , connected by a path @xmath37 in @xmath0
. then @xmath37 must contain two sub - paths in @xmath0 , say @xmath229 for @xmath230 , connecting @xmath231 to @xmath232 and with @xmath233 for all @xmath234 .
since @xmath227 and @xmath228 are @xmath201-distant good boxes , by lemma [ l2 ] , these sub - paths must be edge disjoint . by the definition of a good box , @xmath235 and @xmath236 , which yields @xmath237 by repeating this argument inductively ( on the number of good boxes which are crossed by @xmath37 )
one can get .
lemma [ ecr1 ] follows directly from .
together with lemma [ ecr1 ] , lemma [ l3 ] implies theorem [ t1 ] under . for the general case , assume @xmath238 .
fix @xmath83 so that @xmath239 ( we can do so since @xmath3 is right - continuous ) .
define the auxiliary process @xmath240 and denote by @xmath241 the first - passage time associated to the collection @xmath242 .
thus @xmath243 . since @xmath244 has a bernoulli distribution with parameter @xmath245 , together with the previous case this yields theorem [ t1 ] .
this work was develop during my doctoral studies at impa and i would like to thank my adviser , prof .
vladas sidoravicius , for his dedication and encouragement during this period .
i also thank the whole administrative staff of impa for their assistance and cnpq for financing my doctoral studies , without which this work would have not been possible .
vahidi - asl , m.q .
; wierman , j.c .
first - passage percolation on the voronoi tessellation and delaunay triangulation , random graphs 87 ( m. karonske , j. jaworski and a. rucinski , eds . ) , 1990 , 341 - 359 .
vahidi - asl , m.c . ; wierman , j.c . ( 1992 ) .
a shape result for first - passage percolation on the voronoi tessellation and delaunay triangulation , random graphs 89 ( a. frieze and t. luczak , eds . ) , wiley , 247 - 262 | we consider a first - passage percolation ( fpp ) model on a delaunay triangulation @xmath0 of the plane . in this model
each edge @xmath1 of @xmath0 is independently equipped with a nonnegative random variable @xmath2 , with distribution function @xmath3 , which is interpreted as the time it takes to traverse the edge .
vahidi - asl and wierman @xcite have shown that , under a suitable moment condition on @xmath3 , the minimum time taken to reach a point @xmath4 from the origin @xmath5 is asymptotically @xmath6 , where @xmath7 is a nonnegative finite constant . however
the exact value of the time constant @xmath7 still a fundamental problem in percolation theory . here
we prove that if @xmath8 then @xmath9 , where @xmath10 is a critical probability for bond percolation on the dual graph @xmath11 . |
recent results from a number of independent experiments @xcite can be interpreted as due to finite neutrino masses and , in particular , high statistics measurements of atmospheric neutrinos by the super - kamiokande@xcite experiment are regarded by most as firm evidence that neutrinos have non - zero masses .
+ while these measurements , based on oscillations , have set the stage for a systematic study of the intrinsic neutrino properties , only upper - limits exist on the absolute magnitude of neutrino masses .
indeed , theoretical models span a large range of scenarios , from the degenerate case where mass differences among flavors are small with respect to the absolute mass scale @xcite , to the hierarchical , where mass differences are of the same order as the mass themselves .
however , while the neutrino mass scale is unknown , the present data on oscillations lead rather naturally to masses in the range @xmath3 ev , as shown recently , e.g. in @xcite .
+ it is unlikely that direct neutrino mass measurements , most notably with tritium @xcite , will be able to reach sensitivities substantially below 1 ev in the near future .
in contrast , we will show that a large double-@xmath4 decay experiment using isotopically enriched @xmath5can reach a sensitivity corresponding to neutrino masses as low as @xmath00.01 ev .
a xenon detector offers the unique possibility of identifying the final state , thus providing an essentially background - free measurement of unprecedented sensitivity for the neutrino mass .
it is well known that neutrinoless double beta decay , @xmath6 , can proceed only if neutrinos are massive majorana particles . if the @xmath6occurs , the _ effective _ majorana neutrino mass @xmath7 is related to the half - life @xmath8 as : @xmath9 where @xmath10 is a known phase space factor depending on the end - point energy @xmath11 and the nuclear charge @xmath12 , and @xmath13 and @xmath14 are the gamow - teller and fermi nuclear matrix elements for the process . here we have defined @xmath15 with @xmath16 being the mixing matrix in the lepton sector , and @xmath17 the masses of the individual majorana neutrinos .
hence , although difficulties in the nuclear models used to calculate the matrix elements give some uncertainty on the value of @xmath7 ( see , e.g. @xcite ) , @xmath6decay is sensitive to the masses of all neutrino flavors , provided that the mixing angles are non - negligible .
at present the best sensitivities to @xmath6decay have been reached with @xmath18 diode ionization counters with an exposure of 24.16 kg yr @xcite and with a @xmath2 time projection chamber ( tpc ) with an exposure of 4.87 kg yr @xcite .
the measured half - life limits of @xmath19 yrs for germanium and @xmath20 yrs for xenon can be interpreted as neutrino mass limits of , respectively , @xmath21 ev and @xmath22 ev using the quasi - particle random phase approximation ( qrpa ) @xcite ( shell model ( nsm ) @xcite ) for the nuclear matrix element calculation .
the germanium detector rejects background on the basis of its excellent energy resolution and of pulse - shape discrimination @xcite and has recently reported a specific background as low as 0.3 kg@xmath23 yr@xmath23 fwhm@xmath23 . the relatively inferior energy resolution in tpcs
have been complemented by superior tracking capabilities that allow the xenon experiment to partially reconstruct the two - electron topology of @xmath24-decay , obtaining a specific background of 2.5 kg@xmath23 yr@xmath23 fwhm@xmath23 @xcite .
none of these backgrounds is sufficient for a decisive experiment in the interesting region discussed above .
for that , it is essential to find a reliable way to drastically increase the size of the detectors while , simultaneously , reducing the backgrounds .
this dual requirement stems from the fact that in a background - free experiment the neutrino mass sensitivity scales as @xmath25 , where @xmath26 is the measurement time and @xmath27 the number of nuclei under study . in the opposite extreme ,
when the background scales with @xmath28 , the neutrino mass sensitivity would scale only as @xmath29 .
obviously , the backgrounds observed in all current experiments will become the true limiting factor of any kind of future large experiment , hampering the full utilization of very large masses of @xmath24 emitters .
unlike other isotopes , however , @xmath5allows for direct tagging of the ba - ion final state using optical spectroscopy , as pointed - out for the first time in @xcite .
while this technique can not discriminate between @xmath6and @xmath30-decays , this second process is not the dominant background at the sensitivities sought here and can be eliminated by kinematical reconstruction in a xenon tpc .
moreover , xenon is one of the easiest isotopes to enrich , and like argon can be used as active medium in ionization chambers .
hence , an experiment with xenon can use an entirely new variable in order to drastically reduce the backgrounds and explore the range of interesting neutrino masses .
the @xmath24-decay of @xmath5produces a ba@xmath31 ion in the final state that can be readily neutralized to ba@xmath32 by an appropriate secondary gas in the tpc .
the barium ion can then be individually detected through its laser - induced fluorescence .
single ba@xmath32 ions were first observed in 1978 @xcite using a radio frequency quadrupole trap and laser cooling .
+ the level structure of the alkali - like ba@xmath32 ion is shown in figure [ fig : ba_levels ] . due to
the strong 493 nm allowed transition ground - state ions can be optically excited to the @xmath33 state from where they have substantial branching ratio ( 30% ) to decay into the meta - stable @xmath34 state .
specific ba@xmath32 detection is then achieved by exciting the system back into the @xmath33 state with 650 nm radiation and observing the blue photon from the decay to the ground state ( 70% branching ratio ) .
this transition has a spontaneous lifetime of 8 ns and when saturated will radiate @xmath35 photons / s .
a pair of lasers tuned onto the appropriate frequencies and simultaneously steered to the place where the @xmath24-decay candidate event is found can provide a very selective validation , effectively providing a new independent constraint to be used in @xmath24-decay background subtraction .
the light from the p to s transition conveniently lays in the region of maximum quantum efficiency of bialkali photocathodes so that an array of conventional large - area photomultipliers can be used for the detection .
the very large saturation rate makes the experiment possible even with modest photocathode angular coverage . while it is possible in principle to steer the lasers anywhere inside the tpc , the very large volume and the need for light baffling may favor a scheme in which the ba@xmath32 drift in the large electric field is used to bring the ion in specific laser detection regions . the primary difference between the single atom work done previously and this experiment is that here the @xmath1 ion is not in vacuum but rather in a buffer gas ( xe ) at a pressure of several atmospheres .
the high pressure xenon has two effects : first , it effectively traps the barium since diffusion in the dense gas is sufficient to confine the atoms for long enough time to obtain a signal ; second , it pressure broadens the optical transitions increasing the laser power required . we calculate that at 5 atm the barium ion will diffuse only @xmath36 mm in 1 s , and during this time it can be cycled over @xmath37 times through the three - level scheme of figure [ fig : ba_levels ] , emitting more than @xmath37 493 nm photons .
the ion drift in the electric field of the tpc can be accurately measured and corrected for in the process of steering the lasers . for the moment we notice that drift velocities of tl@xmath32 in xe have been measured @xcite and are low enough to make the correction possible .
we also find from preliminary calculations that at the same pressure the broadened line - width is @xmath38 ghz , 1000 times greater than the natural line - width .
this results in a saturation intensity of @xmath39 w/@xmath40 .
ba@xmath32 lifetime , together with pressure broadening of the lines involved , and the lifetime of the @xmath34 state , will have to be measured in xenon before the detection system can be fully optimized .
however , measurements on the corresponding neutral ba states at 0.5 atm in he and ar @xcite show that operations at 5 to 10 atm should be possible .
among the different gas ionization detectors known , a tpc is the ideal detector for @xmath24-decay as it has no wires or other materials in the detection volume hence reducing the background internal to the detector and simplifying the laser scanning .
the requirements of size , efficiency for contained events , energy , and spatial resolution can be achieved with a large ( 40 m@xmath41 ) gas - phase tpc with xenon .
tpcs of similar size at atmospheric pressure are successfully in operation @xcite , in one case with some 10 kg of @xmath5 @xcite . a 5 to 10 atm
, 40 m@xmath41 chamber would contain 1 to 2 tons of @xmath5and is within the reach of current state - of - the - art technology .
the ultimate sensitivity of the technique is probably limited by the practical availability of @xmath5that represents 8.9% of natural xenon and has to be extracted by isotopic separation . while the isotopic separation is indeed one of the main challenges of the experiment , in the case of @xmath5this operation is simplified by the fact that xenon is a gas at standard conditions , and essentially any known separation technology can be used .
relatively large amounts of @xmath5have been obtained in the past by ultracentrifugation @xcite with purities sufficient for the experiment described @xcite .
@xmath5handling is simplified by its inert character that allows for the extraction of cosmogenically produced elements with distillation and chemical filtering that can be carried - on in the underground laboratory prior ( and during ) the operation of the experiment .
an experiment with up to 10 tons of @xmath5is consistent with the availability of xenon on the world market @xcite and can be setup by assembling a few tpc modules inside a single pressure vessel or , possibly , by increasing the pressure and/or volume of a single chamber .
the use of liquid - xenon ( lxe ) in a tpc would result in a very compact detector with considerable advantages .
however the range of 1.2 mev electrons from the @xmath6-decay in lxe is only of 2.4 mm so that , given any conceivable spatial resolution , the topological information , essential for background rejection , would be lost .
in contrast , at room temperature 5 atm of xenon gas correspond to a density of 30
g / l and an electron range of 21.6 cm at 1.2 mev . in figure
[ fig : tpc_mech ] we show a possible scheme for the @xmath24-decay detector . + it is known @xcite that it is quite difficult to obtain electron multiplication and stable operation in xenon chambers .
this is due to the far uv scintillation light ( @xmath42 nm corresponding to 6.93 ev ) , copiously produced with the multiplication process , that in turn ionizes the gas and extracts electrons from the metallic surfaces . in our case
this particular requirement is greatly relaxed by the use of gas micropattern detector @xcite planes instead of a more conventional anode wire array for the chamber readout , drastically limiting the solid angle available for the uv radiation @xcite .
in addition , the chamber will also use a quencher gas , able to absorb the xenon scintillation uv and re - emit light in the blue or green so that timing information can be recovered using the photomultipliers and the event can be localized also in the third ( time ) coordinate .
a field of @xmath01 kv / cm will be needed in order to achieve high drift velocity for electrons in xenon . while this field strength is not sufficient to neutralize ba@xmath31 to ba@xmath32 , an appropriate choice of ionization potential ( ip ) for the quencher gas can achieve this purpose .
ips between 5.2 ev ( ba ) and 10.001 ev ( ba@xmath32 ) are found in several organic molecules that would provide a stable environment for ba@xmath32 .
molecules like tma ( @xmath43 ) , tea ( @xmath44 ) , tmg ( @xmath45 ) and tms ( @xmath46 ) are candidates fulfilling the above requirements .
an underground site with an overburden of @xmath47 mwe ( meter water equivalent ) will reduce the cosmic ray muon flux to @xmath48 , corresponding to less than 0.1 hz through the detector .
muons recorded in the tpc have a distinctive signature and indeed were observed and rejected by the gotthard group @xcite based on track length , lack of scattering ( high energy ) , and specific ionization .
the last two features provide good discrimination even for tracks that clip a small corner of the chamber .
the online trigger processor will be able to analyze and reject most muon tracks without activating the laser ba - tagging system . based on simple geometrical considerations we expect the laser system to be activated by muon tracks less than once per hour .
some background will be produced by neutrons produced by muon spallation in the structures ( rock and other materials ) outside the detector .
these fast neutrons enter the detector with some efficiency and produce spallation reactions on the xenon , carbon and hydrogen nuclei contained in the tpc . while the original muon and the neutron trail go undetected , the spallation processes provide very distinct high ionization , short tracks and can be distinguished and rejected even at trigger level . hence at the depth considered the detector does not need an active veto counter , as already found by the gotthard group . + more serious are the @xmath49-ray backgrounds from natural radioactivity , either produced outside the tpc ( mainly by the rock or concrete ) , or inside ( mainly @xmath50rn and , possibly , @xmath51kr and @xmath52ar ) .
external @xmath49-ray backgrounds from the rock and concrete can be attenuated by a @xmath025 cm thick lead or steel enclosure and by the 1 cm thick pressure vessel that will be built out of low - activity steel .
additionally , cleaner shielding could be provided , if needed . in the following
we conservatively assume that the tpc itself ( without ba - tagging system ) will have the same specific rate of mis - identified background as the gotthard experiment , and we rely on ba - tagging for the final step of reduction .
this hypothesis is conservative since the larger volume of our tpc provides better self shielding for the radiation produced externally .
we estimate the position and energy resolution of our chamber from an extrapolation using the gotthard tpc . assuming a drift velocity similar to the one in gotthard and a drift distance of 250 cm ( 3.6 times larger than the gotthard case ) we obtain transverse and longitudinal position resolutions ( @xmath53 ) of better than 5 mm .
this figure should give us roughly the same background suppression factor as obtained at gotthard , since the electron range is 21 cm .
furthermore , unlike in the case of the gotthard experiment , the knowledge of the time of occurrence will provide longitudinal localization of each event , improving the understanding of external backgrounds and allowing for a longitudinal fiducial volume cut .
+ energy resolution should be no worse than the @xmath54 ( at 1.592 mev ) obtained at gotthard , since in addition to the total charge , also the scintillation light will be collected @xcite .
the possibility of using the laser tagging system with a small spot - size in a `` raster scan '' mode around the event location , will localize the decay vertex with mm precision and allow a full kinematical fit including energy , angular correlation and range for _ each _ of the electrons .
it is expected that substantial improvement in energy resolution and background rejection will be possible in this way .
+ while the quantitative advantage of each of the techniques discussed will be better understood after extensive laboratory tests and a full monte carlo simulation , here we simply assume that the above methods , together with the ba@xmath32 identification , will reduce the sum of all radioactivity backgrounds by at least three orders of magnitude with respect to the gotthard case , making a 10-ton @xmath5experiment essentially background free .
from the size and geometry of the chamber we obtain an efficiency for fully contained events of 70% . in table
[ tab : performance ] we compare the projected sensitivity of this experiment with the present limits on @xmath6decay .
.comparison of the best present double - beta decay experiments and the project described here .
the technique of ba@xmath32 tagging will reduce the background , given in units of kg@xmath23 yr@xmath23 fwhm@xmath23 if not noted otherwise , to the level necessary to fully utilize the large mass of isotopic species .
all other experimental parameters are assumed to be the same as in the gotthard experiment .
we list here , together , the case of an initial detector with 1 ton of @xmath5and the final results possible with 10 tons of @xmath5and a very long ( 10 years ) data - taking period .
the quantities marked with @xmath55 are radioactivity backgrounds that are assumed to be negligible as discussed in the text .
in addition the background from mis - identified @xmath30decays is also shown in total events in each exposure using the asymmetric analysis interval described in the text . [ cols="<,^,^,^,^,^,^,^,^ " , ] it is interesting to note that for the very large isotopic masses contemplated here , the background on @xmath6decay from the well known @xmath30mode has to be carefully estimated .
as already remarked this `` background '' is not directly suppressed by the laser tagging methods . however , the energy spectra of the electrons that represent the only distinctive feature of this background will be better measured thanks to the better knowledge of the event kinematics . in the following
we will conservatively disregard this additional information and suppress the @xmath30mode using total energy information , with the resolution mentioned above . in order to understand the role of the resolution we select events in the two intervals @xmath56 $ ] and @xmath57 $ ] .
we then compute the number of @xmath30decay events @xcite left in each case .
for @xmath58 we have 23 @xmath59 left in @xmath60 and 0.36 @xmath59 left in @xmath61 . a better resolution of @xmath62 is used in the 10 ton - 10 yr exposure in table [ tab : performance ] . the loss in efficiency due to the asymmetric cut ( and to the tails beyond @xmath63 ) are taken into account in the table as appropriate .
it is clear , however , that our estimate is quite conservative as the @xmath30decay background will be further suppressed by the kinematic fitting .
the projected neutrino mass sensitivity of 10 - 50 mev would make the discussed experiment competitive with other large scale double beta searches which have been proposed .
nemo3 is scheduled to start data taking in 2001 with initially 7 kg of @xmath64mo and 1 kg of @xmath65se in form of passive source foils in a gas tracking detector utilizing a magnetic field and a scintillator calorimeter @xcite .
a neutrino mass sensitivity of 0.1 ev is expected after 5 years .
the ultimate goal is to run with 20 kg of @xmath65se , @xmath64mo or @xmath66nd in order to measure an effective neutrino mass below 0.1 ev .
a large cryogenic detector ( cuore ) , able to operate 600 kg of teo@xmath67 crystals , could be used to search for the @xmath6decay of @xmath68
. a neutrino mass sensitivity of around 0.1 ev has been quoted for this device @xcite .
the proponents of the genius project propose to use up to one ton of isotopically enriched @xmath69ge suspended in a large tank of liquid nitrogen @xcite .
the goal is to reach a mass sensitivity of around 0.01 ev .
the latter two are calorimetric approaches .
however , it has to be pointed out that the experiment discussed in this paper is the only one among those third generation projects which plans to utilize a novel technique for background suppression in the form of ba tagging . in summary ,
we have described an advanced @xmath24 detector system that uses a hybrid of atomic and particle physics techniques to qualitatively improve background suppression .
such detector opens new possibilities in using massive amounts of @xmath5for an advanced @xmath24-decay experiment that would explore neutrino masses in the range 10 - 50 mev , providing a unique opportunity for discoveries in particle physics and cosmology .
we would like to specially thank f. boehm for the help and guidance received in formulating much of this paper .
we also owe gratitude to u. becker , s. chu , h. henrikson , l. ropelewski , t. thurston and r. zare for many useful discussions .
finally we would like to thank r.g.h .
robertson for pointing out an inconsistency in the original version of table [ tab : performance ] . | we describe an approach to the study of neutrino masses that combines quantum optics techniques with radiation detectors to obtain unprecedented sensitivity .
with it the search for majorana neutrino masses down to @xmath010 mev will become accessible .
the experimental technique uses the possibility of individually detecting @xmath1-ions in the final state of @xmath2 double - beta decay via resonant excitation with a set of lasers aimed at a specific location in a large time projection chamber .
the specificity of the atomic levels provides tagging and , together with more traditional event recognition parameters , greatly suppresses radioactive backgrounds .
double beta decay , neutrino mass , tpc , optical tagging , ion trap |
it has been widely known that properties of hadrons undergo changes at finite temperature and/or density . in particular , the modification of vector mesons in medium plays an essential role in understanding the properties of the qcd vacuum under hot and/or dense environment ( see , for example , the following references and those therein @xcite ) , since there may be a possible connection between their in - medium modification and the restoration of chiral symmetry of which spontaneous breakdown provides a mechanism for the mass generation of most hadrons . in particular , it is of great interest to see the in - medium change of the @xmath0 meson , since it is strongly coupled to the @xmath3 isovector channel and may be observed in dilepton productions ( @xmath4 ) in heavy - ion collision because of the fact that the dileptons produced from the @xmath0 meson can convey information on the mass shift of the @xmath0 meson without any contamination due to the strong interaction .
in fact , there are already experimental results for changes of the @xmath0-meson properties in medium : the ceres / na45 collaboration has investigated low - mass electron pair productions in various collisions @xcite .
the na60 collaboration has recently announced that the averaged @xmath0-meson spectral function shows a strong broadening but no shift in mass @xcite .
the clas collaboration has reported that the medium effects on the @xmath0-meson mass are not observed @xcite and some widening of the @xmath0-meson decay width is seen for the fe - ti target .
however , the e325 experiment at the kek 12-gev proton synchrotron has measured the invariant mass spectra of @xmath5 pairs in @xmath6 reactions and has found that the @xmath0-meson mass is dropped by about @xmath7 @xcite .
the broadening of the @xmath0-meson width does not seem to fit the e325 data .
the star collaboration has also reported a possible shift of the @xmath0-meson peak in the @xmath8 channel in au+au and @xmath9+@xmath9 collisions at @xmath10 @xcite . the tagx collaboration has observed an in - medium modification of the @xmath0-meson spectral function in @xmath11 , @xmath12 , and @xmath13 reactions @xcite .
the in - medium change of the @xmath0 meson has been extensively studied in many different theoretical approaches .
brown and rho @xcite first proposed a possible dropping of the @xmath0-meson mass , introducing the effective scalar glueball field in order to make the effective chiral lagrangian consistent with the qcd scaling property . ref .
@xcite estimated the @xmath0-meson mass in medium dropped by about @xmath14 .
subsequently , hatsuda and lee have shown its mass shifted by about @xmath15 @xcite , based on the qcd sum rules .
carter et al .
@xcite , on the contrary , have found that it turns out much smaller , i.e. around @xmath16 , since the glueball field is rather stable in medium in their effective lagrangian approach .
klingl et al .
@xcite , however , have investigated the in - medium @xmath0-meson properties in the qcd sum rules but have concluded that the change of the @xmath0 meson in medium can not be interpreted simply by the mass dropping of the @xmath0 meson alone . refs .
@xcite have shown in the study of the @xmath0-meson spectral function that the width broadening is more important to interpret the corresponding experimental data .
thus , no consensus has been reached , either experimentally or theoretically , as yet on how the @xmath0 meson experiences the change in medium . in the present work ,
we want to investigate the electromagnetic charge radii of the pion at finte density but @xmath17 within the framework of the nonlocal chiral quark model ( n@xmath18qm ) from the instanton vacuum , based on which the mass shift of the @xmath0 meson will be evaluated , associating with the vector meson dominance ( vmd ) in a phenomenological approach .
we already studied the electromagnetic ( em ) form factor of the pion within the same framework @xcite . since the instanton vacuum at finite density was investigated in ref .
@xcite some years ago , we can immediately use the formalism of ref .
@xcite to extend our previous work into finite density .
thus , we first consider in this work the pion em form factor at finite quark chemical potentials ( @xmath19 ) and its charge radii .
having shown how much the pion charge radius is modified in medium , we try to translate its change at finite density into the corresponding shift of the @xmath0-meson mass in medium with the help of the vmd .
the present work is organized as follows : in section ii , we make a brief introduction of the n@xmath18qm and its application to the nonzero @xmath19 .
the parameterized pion em form factor in terms of the vmd is introduced in order to determine the mass dropping of the @xmath0 meson .
the numerical results with discussions are given in section iii .
the final section is devoted to the summary and conclusion .
we start with the definition of the pion em form factor in the space - like region : @xmath20 where @xmath21 stands for the positively charged pion state ( @xmath22 ) .
the @xmath23 and @xmath24 represent the initial and final on - shell momenta for the pion , satisfying @xmath25 .
the pion mass @xmath26 is taken to be 140 mev for numerical input .
the momentum transfer is defined as @xmath27 and @xmath28 in the space - like region . here , the local em current in flavor su(2 ) is defined in euclidean space as follows : @xmath29 with the charge operator @xmath30 .
note that all calculations are performed in euclidean space , since we are interested in the form factor in the space - like region .
the em form factor satisfies the normalization condition at @xmath31 by charge conservation : @xmath32 . in ref .
@xcite , we have already computed the pion em form factor in free space , employing the nonlocal chiral quark model ( n@xmath18qm ) from the instanton vacuum @xcite .
a great virtue of the n@xmath18qm from the instanton vacuum lies in the fact that there are only two parameters , namely , the average instanton size @xmath33 and average inter - instanton distance @xmath34 .
the normalization scale of this approach can be defined by the average size of instantons and is approximately equal to @xmath35 mev .
the values of the @xmath36 and @xmath37 were estimated phenomenologically in ref .
@xcite as well as theoretically in ref .
furthermore , it was confirmed by various lattice simulations of the qcd vacuum @xcite .
moreover , lattice calculations of the quark propagator @xcite are in a remarkable agreement with that of ref . @xcite . a recent lattice simulation with the interacting instanton liquid model obtains @xmath38 and @xmath39 with the finite current quark mass @xmath40 taken into account @xcite .
the n@xmath18qm turns out to be very successful in describing properties of hadrons in the low - energy regime .
however , the presence of the nonlocal interaction between quarks and pseudo - goldstone bosons breaks the gauge invariance for the external vector fields .
since the pion em form factor involves the vector current , we need to deal with this problem . while ref .
@xcite proposed a systematic way as to how the conservation of the nther current is restored , one has to handle the integral equation .
@xcite derived the light - quark partition function in the presence of the external gauge fields . with this gauged partition function
, it was shown that the low - energy theorem for the transition from two - photon state to the vacuum via the axial anomaly was satisfied @xcite . moreover
, the gauged effective chiral action has been shown to describe well a great deal of mesonic properties such as the meson distribution amplitudes and kaon semileptonic decay form factors @xcite .
as shown in refs .
@xcite , if the external vector fields are weak enough , the gauged effective chiral action in the chiral limit can be written in terms of the covariant derivative @xmath41 : @xmath42 , \end{aligned}\ ] ] where the functional trace @xmath43 runs over space - time , color , flavor and dirac spaces .
the @xmath44 is the covariant derivative expressed as @xmath45 , where @xmath46 stands for the external em gauge field .
the @xmath47 denotes the nonlinear goldstone boson fields defined as : @xmath48 where @xmath49 stands for the pion decay constant : @xmath50 mev .
note that the pion em form factor derived from eq .
( [ eq : eca ] ) can be shown to satisfy obviously the ward - takahashi identity .
we refer to ref .
@xcite for the detailed formalism of how to derive the pion em form factor from eq .
( [ eq : eca ] ) .
now , we are in a position to introduce the quark chemical potential ( @xmath19 ) in the model in order to calculate the pion em form factor in medium . since we consider the case of the chiral limit , isospin symmetry is automatically taken into account : @xmath51 .
thus , we need to consider only the isoscalar quark chemical potential . in the present work ,
we closely follow ref .
@xcite . moreover , we take into account the gauge invariance in the presence of the external em field , as mentioned before .
the relevant effective chiral action at finite density can be written as follows : @xmath52=-{\rm sp}\ln\left[i\rlap{/}{d}- i\rlap{/}{\mu } + i\sqrt{{\cal m}(id ) } u^{\gamma_5}\sqrt{{\cal m}(id ) } \right ] , \end{aligned}\ ] ] where @xmath53 denotes the quark chemical potential as a four vector .
the momentum - dependent quark mass , @xmath54 , modified by the nonzero quark chemical potential , can be obtained by the fourier transform of the following modified dirac equation for the quark zero mode in the instanton vacuum : @xmath55\psi^{(0)}_{i\bar{i } } = 0.\end{aligned}\ ] ] the momentum - dependent quark mass at finite density has been obtained already in ref .
@xcite . however , it is rather complicated to use that derived in ref .
@xcite , so that in the present work we make a parameterization for @xmath56 in the following simple form : @xmath57 ^ 2={\cal m}_0{\cal f}^2(i\partial,\mu),\end{aligned}\ ] ] where @xmath58 denotes the quark form factor which is related to eq .
( [ eq : mde ] ) .
note that there is a caveat in using eq .
( [ eq : mfd ] ) : such a simple parameterization is only valid at dilute density ( @xmath59 mev ) @xcite .
otherwise , the results would become unstable . in fig . 1
, we compare the parameterized quark form factor given in eq .
( [ eq : mfd ] ) ( solid curves ) with that from ref .
@xcite ( dashed curve ) at @xmath60 mev and @xmath61 mev . in eq .
( [ eq : mfd ] ) with that of ref .
the solid curve depicts the parameterized form factor , while the dashed one draws that of ref .
the form factors are calculated at @xmath60 mev and @xmath61 mev.,title="fig:",width=377 ] [ fig:1 ] as shown in fig . 1
, the simple parameterized form factor reproduces that of ref .
@xcite qualitatively well .
the cut - off mass @xmath62 , which is taken to be the inverse of the avergage instanton size , can be consistently determined by the self - consistency equation : @xmath63 although there must be corrections due to the presence of the quark chemical potential , they are very small as far as we stay below the region below normal nuclear matter density ( @xmath64 mev ) .
hence , we take the vacuum value @xmath65 and it will give @xmath66 mev .
following the same formalism as in ref .
@xcite and introducing the nonzero quark chemical potential , we obtain the following local and nonlocal contributions to the pion em form factor in eq .
( [ eq : emff ] ) at @xmath67 : @xmath68 , \nonumber\\ \label{eq : nonlocal } & & f^{*\ , \mathrm{nonlocal}}_{\pi } = \sum_{\rm flavor } \frac{8e_qn_c}{(2p_i\cdot q+m^2_{\pi } ) } \int^{\infty}_{-\infty}\frac{d^4k}{(2\pi)^4}\bigg [ \frac{\sqrt{{\cal m}_b{\cal m}_c}(\sqrt{{\cal m}_c}\hat{\cal m}_{bd}-\sqrt{m_b}\hat{\cal m}_{cd } ) ( k_{bc}-{\cal m}_b{\cal m}_c)}{(k^2_b+{\cal m}^2_b)(k^2_c+{\cal m}^2_c ) } \nonumber\\ & & \hspace{1cm}+\frac{{\cal m}_a\sqrt{{\cal m}_b{\cal m}_c}(\sqrt{{\cal m}_b } \hat{\cal m}_{cd}-\sqrt{{\cal m}_c}\hat{\cal m}_{bd})({\cal m}_ck_{ab } + { \cal m}_bk_{ac}-{\cal m}_ak_{bc}+{\cal m}_a{\cal m}_b{\cal m}_c ) } { ( k^2_a+{\cal m}^2_a)(k^2_b+{\cal m}^2_b)(k^2_c+{\cal m}^2_c ) } \nonumber\\ & & \hspace{1cm}+\frac{\sqrt{{\cal m}_a{\cal m}_c } \left[\sqrt{{\cal m}_b}\hat{\cal m}_{ad}-\sqrt{{\cal m}_a}\hat{\cal m}_{bd}\right ] ( k_{ac}+{\cal m}_a{\cal m}_c ) } { 2(k^2_a+{\cal m}^2_a)(k^2_c+{\cal m}^2_c ) } \nonumber\\ & & \hspace{1cm}+\frac{\sqrt{{\cal m}_a{\cal m}_b } \left[\sqrt{{\cal m}_c}\hat{\cal m}_{ad}-\sqrt{{\cal m}_a}\hat{\cal m}_{cd}\right ] ( k_{ab}+{\cal m}_a{\cal m}_b ) } { 2(k^2_a+{\cal m}^2_a)(k^2_b+{\cal m}^2_b)}\bigg],\end{aligned}\ ] ] where @xmath69 .
the relevant momenta are defined as @xmath70 , @xmath71 , @xmath72 and @xmath73 .
note that all the momenta depend on @xmath19 .
we use the following abbreviations : @xmath74 and @xmath75 is defined as the derivative of the dynamical quark mass : @xmath76 for more details of evaluating eq .
( [ eq : local ] ) , one can refer to ref .
@xcite .
since the pion em form factor derived from the instanton vacuum is only valid in the space - like region , we need further theoretical consideration in order to connect the present calculation to the change of the @xmath0-meson mass .
it is well known that the vector meson dominance ( vmd ) describes the pion em form factor quantitatively in terms of @xmath0-meson exchange with the help of dispersion theory @xcite . however , in order to apply the vmd , we use a modified expression for the pion em form factor in medium written as follows @xcite : @xmath77 where @xmath78 and @xmath79 denote the mass and full decay width of the @xmath0 meson in medium , respectively .
the parameter @xmath80 encodes the change of the pion em form factor in medium . note that from now on the quantities with asterisk are those in medium .
in fact , the prefactor @xmath81 can be written as @xmath82 where @xmath83 and @xmath84 denote the strong coupling constant for @xmath85 and the photon-@xmath0 meson coupling constant , respectively . in free space ,
the universality relation @xmath86 in the vmd allows us to put @xmath87 equal to @xmath88 so that the pion em form factor may be properly normalized to be @xmath89 at @xmath31 .
moreover , the decay width @xmath90 plays only an important role in the vicinity of @xmath91 , so that we will not consider its change here , since we are interested in the pion em form factor in the space - like region .
the pion em charge radius is defined as follows : @xmath92 using eqs .
( [ eq : emffvmd ] ) and ( [ eq : emcr ] ) , we have the following relation : @xmath93^{1/2}.\end{aligned}\ ] ] thus , the in - medium em charge radius can be written as a function of @xmath19 : @xmath94 ^ 2 \approx0.45\,{\rm fm}^2\times{\cal c}^*(\mu_q ) \left[\frac{m_{\rho}}{m^*_{\rho}(\mu_q)}\right]^2,\end{aligned}\ ] ] note that @xmath95 is replaced by the experimental value @xmath96 @xcite .
thus , the @xmath0-meson mass for the nonzero @xmath19 can be easily obtained as follows : @xmath97^{1/2}.\end{aligned}\ ] ] since we can explicitly compute @xmath98 and @xmath99 in the n@xmath18qm , we can easily relate the change of the pion charge radius to that of the @xmath0-meson mass .
we discuss now the results obtained from the present work .
we first want to mention that there is no room to play with parameters in the n@xmath18qm from the instanton vacuum , since all parameters , in particular , @xmath100 and @xmath62 are fixed already by the saddle - point equation .
thus , we will present almost all the results as functions of the quark chemical potential @xmath19 in the folloinwg .
note that the parameterized form factor in eq .
( [ eq : mfd ] ) is only valid in lower @xmath19 . as @xmath19 is getting close to the value of the phase transition ( @xmath101 mev )
, we find numerical instabilities due to a singular behavior of the parameterized quark form factor .
however , as far as we stay in lower @xmath19 , we are safe from such instabilities . at finite density .
the solid curve draws the pion em form factor at @xmath60 mev , while the dashed curve represents that in free space .
the experimental data in free space are taken from refs .
@xcite.,width=377 ] in fig .
[ fig:2 ] , we first show the results of the pion em form factor at @xmath60 mev in comparison with that in free space . while at higer @xmath102 the pion em form factor seems very similar to that in medium
, we observe that the pion em form factor in the vicinity of @xmath31 becomes larger .
moreover , the slope of the form factor gets steeper near @xmath31 with @xmath19 turned on .
this effect will be clearly seen in calculation of the pion chage radius . as a function of the quark chemical potential @xmath19 .
the solid curve depicts the @xmath103 , while the dashed line represents that in free space , i.e. @xmath104 ( experimental data : @xmath105 @xcite).,width=377 ] in fig .
[ fig:3 ] we draw the numerical results for the @xmath106 as a function of @xmath19 .
it is shown that the pion charge radius in medium increases as @xmath19 increases .
it indicates , as mentioned already , that the pion em form factor gets steeper in the vicinity of @xmath31 as @xmath19 increases .
one can infer from this that the pion charge distribution gets smeared in medium as the matter density gets denser .
in other words , the interaction between the quarks inside the pion lose kinetic energies in medium . as a function of @xmath19 .
the solid curve represents the present result , while the dashed line denotes its value in free space , i.e. @xmath107.,width=377 ] figure [ fig:4 ] depicts the prefactor @xmath108 evaluated at @xmath31 in the n@xmath18qm .
being similar to the pion charge radius , the @xmath99 increases as @xmath19 increases . in order to understand the physical meaning of the @xmath80 ,
let us consider eq .
( [ eq : ratio ] ) in which the @xmath109 is given as a ratio @xmath110 and @xmath111 .
it indicates from the result shown in fig .
[ fig:4 ] that @xmath110 and @xmath112 in medium are modified in different ways , though we can not tell how they are changed explicitly .
-meson mass @xmath113 as a function of @xmath114 .
the baryon chemical potential @xmath115 is defined as @xmath116 .
the @xmath117 denotes the chemical potential corresponding to normal nuclear matter density @xmath118.,width=377 ] finally , we are in a position to discuss the mass shift of the @xmath0 meson . inserting the results of @xmath98 and @xmath99 as already shown in figs .
[ fig:3 ] and [ fig:4 ] , respectively , into eq .
( [ eq : mrho ] ) , we arrive at the final result of the @xmath0-meson mass shift in medium as shown in fig .
[ fig:5 ] . obviously , the @xmath0-meson mass decreases as the @xmath19 increases .
this is due to the fact that the result of the pion charge radius squared , @xmath98 , grows faster than that of the @xmath99 . as a result , the @xmath0-meson mass , which is proportional to a ratio of @xmath119 as shown in eq .
( [ eq : mrho ] ) , falls off approximately linearly as @xmath19 increases . since the @xmath0-meson mass decreases approximately linearly , we are able to employ the following linear parameterization for the mass shift of the @xmath0 meson as in ref .
@xcite : @xmath120 where the @xmath115 denotes the baryon number chemical potential being equal to @xmath121 .
the @xmath117 denotes the chemical potential corresponding to normal nuclear matter density , @xmath122 , and is chosen to be about @xmath123 mev .
we note that @xmath117 was estimated to be @xmath124 mev , the nucleon mass and its binding energy being taken into account @xcite .
@xmath125 is a coefficient for showing the mass dropping of the @xmath0 meson .
in fact , the mass ratio of eq .
( [ eq : para ] ) is usually parameterized as a function of a ratio of the nuclear matter densities @xmath126 @xcite : @xmath127 . however , in the lower @xmath19 region ( @xmath128 mev ) , the baryon number density turns out to be zero due to the following relation @xmath129^{3/2}$ ] in which @xmath130 is the constituent quark mass , @xmath131 mev @xcite .
hence , we only show in the present work the @xmath0-meson mass shift as a function of @xmath114 .
the value of @xmath125 is then determined to be @xmath132 , i.e. @xmath133 . the present result is slightly smaller than refs .
@xcite : in ref .
@xcite , @xmath125 was estimated as @xmath134 , which is similar to that based on phenomenological method in ref .
we want to mention also that , in contrast , the @xmath0-meson mass drops suddenly as the @xmath19 increases in ref .
@xcite using the lattice simulation in color su(2 ) symmetry .
the present result @xmath132 is also consistent with a recent measurement at kek @xcite .
we have investigated the pion electromagnetic form factor and pion charge radius at finite density within the framework of the nonlocal chiral quark model from the instanton vacuum .
the pion electromagnetic form factor is getting steeper near @xmath31 as @xmath19 increases . as a result
, the pion charge radius grows as @xmath19 increases . with the help of the vector meson dominance
, we were able to express the @xmath0-meson mass shift in terms of the prefactor , which is written as a ratio of the @xmath0-@xmath135-@xmath135 strong coupling constant and @xmath0-photon constant , and pion charge radius .
in the present work , we have observed that the pion charge radius increases with respect to the quark chemical potential which implies that the pion charge distribution is smeared in medium as the density gets denser and the interaction between quarks gets lessened .
the prefactor @xmath81 also is getting stronger as @xmath19 increases .
with these two quantities , we have shown that the @xmath0-meson mass falls off as @xmath19 increases , which is similar to that suggested by brown - rho scaling law .
when we made a linear parameterization for the mass shift ( @xmath136 ) , the coefficient @xmath125 turns out to be about @xmath137 , which is compatible to those of other models and phenomenological estimations ( @xmath134 ) .
moreover , the present result is consistent with a recent measurement of the in - medium @xmath0-meson mass shift @xcite .
we finally stress that the mass shift of the @xmath0-meson can be also tested by measuring the pion em charge radius in medium .
thus , we anticipate experiments focussing on the pion em form factor in medium ( @xmath138 and @xmath67 ) to test the present theoretical assertion .
the authors would like to thank t. kunihiro , c. h. lee , s. h. lee , m. harada , m. m. musakhanov , and i. k. yoo for stimulating and fruitful discussions . the work of s.i.n .
is partially supported by the grant for scientific research ( priority area no .
17070002 ) from the ministry of education , culture , science and technology , japan . the work of h.ch.k .
is supported by the korea research foundation grant funded by the korean government(moehrd ) ( krf-2006 - 312-c00507 ) .
the numerical calculations were carried out on yisun at yitp in kyoto university and on miho at rcnp in osaka university .
99 w. cassing and e. l. bratkovskaya , phys .
* 308 * ( 1999 ) 65 .
r. rapp and j. wambach , adv .
* 25 * ( 2000 ) 1 .
g. e. brown and m. rho , phys .
* 398 * ( 2004 ) 301 .
g. agakichiev _ et al . _ [ ceres collaboration ] , phys .
* 75 * ( 1995 ) 1272 .
g. agakichiev _ et al .
_ [ ceres / na45 collaboration ] , phys .
b * 422 * ( 1998 ) 405 .
d. adamova _ et al .
_ [ ceres / na45 collaboration ] , phys .
lett . * 91 * ( 2003 ) 042301 .
r. arnaldi _ et al . _
[ na60 collaboration ] , phys .
* 96 * ( 2006 ) 162302 . c. djalali , r. nasseripour , d. p. weygand and m. h. wood [ clas collaboration ] , j. phys .
g * 34 * ( 2007 ) s495 .
r. nasseripour _ et al .
_ [ clas collaboration ] , arxiv:0707.2324 [ nucl - ex ] .
m. naruki _ et al .
_ , phys .
* 96 * ( 2006 ) 092301 .
j. adams _ et al . _ [ star collaboration ] , phys .
* 92 * ( 2004 ) 092301 .
g. m. huber _ et al . _
[ tagx collaboration ] , phys .
c * 68 * ( 2003 ) 065202 .
g. e. brown and m. rho , phys .
lett . * 66 * ( 1991 ) 2720 .
t. hatsuda and s. h. lee , phys .
c * 46 * ( 1992 ) 34 .
g. w. carter , p. j. ellis and s. rudaz , nucl . phys . a * 603 * ( 1996 ) 367 [ erratum - ibid . a * 608 * ( 1996 ) 514 ] .
f. klingl , n. kaiser and w. weise , nucl .
a * 624 * ( 1997 ) 527 .
r. rapp , g. chanfray and j. wambach , phys .
* 76 * ( 1996 ) 368 .
m. post , s. leupold and u. mosel , nucl .
phys . a * 689 * ( 2001 ) 753 .
s. i. nam and h. -ch .
kim , arxiv:0709.1745 [ hep - ph ] , submitted to prd .
g. w. carter and d. diakonov , phys . rev .
d * 60 * ( 1999 ) 016004 .
d. diakonov and v. y. petrov , nucl .
b * 245 * ( 1984 ) 259 .
d. diakonov and v. y. petrov , nucl .
b * 272 * ( 1986 ) 457 .
d. diakonov , prog .
* 51 * ( 2003 ) 173 .
e. v. shuryak , nucl .
b * 203 * ( 1982 ) 93 .
t. schfer and e. v. shuryak , rev .
* 70 * ( 1998 ) 323 .
m. c. chu , j. m. grandy , s. huang and j. w. negele , phys .
d * 49 * ( 1994 ) 6039 .
j. w. negele , nucl .
* 73 * ( 1999 ) 92 .
t. degrand , phys .
d * 64 * ( 2001 ) 094508 .
faccioli and t. a. degrand , phys .
* 91 * ( 2003 ) 182001 .
o. bowman , u. m. heller , d. b. leinweber , a. g. williams and j. b. zhang , nucl .
proc . suppl .
* 128 * ( 2004 ) 23 .
m. cristoforetti , p. faccioli , m. c. traini and j. w. negele , phys .
d * 75 * ( 2007 ) 034008 .
p. v. pobylitsa , phys .
b * 226 * ( 1989 ) 387 .
m. m. musakhanov and h. -ch .
kim , phys .
b * 572 * ( 2003 ) 181 .
kim , m. musakhanov and m. siddikov , phys .
b * 608 * ( 2005 ) 95 .
s. i. nam and h. ch .
kim , phys .
d * 74 * ( 2006 ) 076005 .
s. i. nam and h. ch .
kim , phys .
d * 75 * ( 2007 ) 094011 .
s. i. nam , h.y .
ryu , m. m. musakhanov , and h .- ch .
kim , in preparation .
j. j. sakurai , `` _ currents and mesons _ '' , the university of chicago press , chicago , 1969 .
r. p. feynman , `` _ photon - hadron interactions _ '' , addison - wesley , reading , massachusetts , 1972 .
c. j. bebek _ et al .
_ , phys .
d * 9 * ( 1974 ) 1229 .
c. j. bebek _ et al .
_ , phys .
d * 13 * ( 1976 ) 25 .
s. r. amendolia _ et al . _
[ na7 collaboration ] , nucl .
b * 277 * ( 1986 ) 168 .
j. volmer _ et al .
_ [ the jefferson lab f(pi ) collaboration ] , phys .
* 86 * ( 2001 ) 1713 .
w. m. yao _ et al .
_ [ particle data group ] , j. phys .
g * 33 * ( 2006 ) 1 .
m. buballa , phys .
* 407 * ( 2005 ) 205 .
s. muroya , a. nakamura and c. nonaka , phys . lett .
b * 551*(2003 ) 305 . | we investigate the pion electromagnetic charge radii and mass dropping of the @xmath0 meson at finite density .
we first calculate the pion charge radii within the framework of the nonlocal chiral quark model from the instanton vacuum both at zero and finite densities . in order to relate the change of the pion charge radius to that of the @xmath0-meson mass at finite density
, we employ the vector - meson dominance for the pion electromagnetic form factor .
it turns out that the pion charge radius is getting larger as the quark chemical potential increases . as a result
, the @xmath0-meson mass falls off as the quark chemical potential grows and is found to be dropped by about @xmath1 at normal nuclear matter density : @xmath2 , which is compatible to the results of the measurement at the kek recently . |
active galactic nuclei ( agn ) provide the majority of photons capable of ionising intergalactic hydrogen over most of the age of the universe ( e.g. * ? ? ?
* ; * ? ? ?
some of the most commonly observed metal absorption lines in the intergalactic medium ( igm ) , including @xmath8 and @xmath5 , have ionisation potentials at extreme uv ( euv ) energies where their main photo - ionising sources are agn .
the mean free paths of ionising photons with energies greater than 1 rydberg are long enough to warrant the assumption that the ionising extra - galactic background ( egb ) is uniform in its spatial distribution ( e.g. * ? ? ?
the main criterion for this assumption is that the mean free paths of photons are longer than the clustering length of photo - ionising sources ( e.g. * ? ? ?
this assumption obviously breaks down near sources of ionising radiation .
particularly notable are the proximity zones of qsos , where the locally generated ionising radiation at euv energies may exceed the egb by orders of magnitude .
the simple nature of hydrogen , having only two ionisation states , allows an elegant relationship where , in ionisation equilibrium , the neutral fraction scales in inverse linear proportion to the ionisation field strength if hydrogen is photo - ionised . a dramatic example of this
are the proximity zones of high-@xmath15 quasars , where the @xmath16 forest absorption is significantly reduced by the enhanced agn radiation field as the qso is approached ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
qso pairs can probe the transverse proximity effect in which the background qso line of sight intersects the foreground qso proximity zone , where the radiation field is increased by orders of magnitude compared to the typical egb ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
metal absorption also reflects the effects of increased ionisation in quasar proximity zones .
proximate absorbers are often defined as being within 5000 @xmath17 of the qso velocity along the line of sight , and show some distinct signatures such as @xmath5 absorption with little or no @xmath2 absorption ( e.g. * ? ? ?
@xcite observe unique metal - line systems at @xmath18 in transverse proximate zones with an ionisation enhancement of 10 - 200@xmath19 , where @xmath5 is enhanced and @xmath8 is weakened .
these signatures suggest that enhanced photo - ionisation affects metal lines , but in a different way than hydrogen .
the multiple ionisation levels of heavy elements makes the response to increased ionisation more complex : metal ion fractions of order unity are possible at diffuse igm densities ( i.e. much higher than @xmath2 fractions ) ; plus these fractions are not a monotonic function of density or temperature like that of @xmath2 , but instead peak at a preferred density or temperature for an intermediate ionisation state . in the case of an increased ionisation field ,
the peak ionisation fraction for a metal species will move in linear proportion with the photo - ionisation rate to higher density if the species is photo - ionised and in equilibrium .
another difference between hydrogen and metal species is the timescale that it takes each to achieve ionisation equilibrium when a proximate agn turns on and off .
the ionisation timescales , relevant for the agn turn on , are longer for metal species with higher ionisation potentials , because fewer ionising photons are produced by the agn at these higher energies . however , the more important difference is the longer timescales metals take to reach ionisation equilibrium after the agn turns off .
although most metal species have shorter recombination timescales , @xmath20 , than @xmath21 , the relevant timescale to achieve equilibrium is @xmath20 times the equilibrium ionisation fraction of the recombined species .
for example , at typical warm ( @xmath13 k ) igm densities of the @xmath16 forest ( @xmath22 ) , @xmath23 yr , but the @xmath2 ionisation fraction , @xmath24 , is @xmath25 ; hence the equilibration timescale is only @xmath26 yr , because hydrogen only has to recombine to a very low neutral fraction to achieve equilibrium .
metal - line species like @xmath8 and @xmath5 have recombination timescales that are @xmath27 yr at similar warm igm densities , but typical ionisation fractions much closer to unity .
therefore , timescales to recombine to equilibrium are also @xmath27 yr much longer than for @xmath2 .
furthermore , the existence of multiple ionisation states for metals means that there are a series of ions to recombine through , and hence multiple recombination timescales , further extending the total equilibration timescale .
agn proximity zone fossils are defined as proximity zones in which the agn has turned off , but ionisation equilibrium has not yet been achieved .
the fossil lifetime can last much longer than the agn - on phase , which is often calculated to be @xmath28 yr ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in this work , we use the non - equilibrium code introduced in @xcite to follow the ionisation and temperature evolution of metal - enriched gas subjected to a variable ionisation field owing to an agn .
we consider several situations where the fossil stage lasts as long as , and sometimes much longer than the agn phase .
we predict the absorption line signatures of these fossil zones , and compare them to some recent observational results .
we focus on diffuse gas with hydrogen numbers densities @xmath0 ( i.e. @xmath29 ) , initially at the equilibrium temperature where cooling balances photo - heating , corresponding to warm igm temperatures , @xmath30 k. the first case we will consider may be applicable to the case of the recently discovered strong @xmath6 absorber detected by @xcite in the cosmic origins spectrograph ( cos ) spectrum of pg1206 + 459 at @xmath31 .
this system is one of the most unique and most puzzling metal absorbers ever observed , containing both high ions like @xmath6 and low ions like @xmath10 .
we explore non - equilibrium photo - ionised models with the agn on and in the fossil phase .
we bolster our argument with evidence for a post - starburst , @xmath32 galaxy , possibly within 100 proper kpc , which shows signatures of being an agn .
we use this as a fiducial example to consider variations in agn strength , lifetime , and duty cycles to show how metal lines can respond to variable ionisation conditions .
we next consider agn proximity zone fossils at the height of qso activity at @xmath33 .
the observations of metal - line absorption in transverse qso proximity zones along the line of sight ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) provide our initial motivation , and we extend the exploration to the case where the proximity zones turn into fossil zones after the foreground qsos turn off .
we argue that unique absorption signatures should occur where @xmath2 re - equilibrates , but @xmath5 and @xmath8 are still out of equilibrium .
we suggest that during the agn era a large fraction of metal - line systems could reside in proximity zone fossils .
the last example we consider concerns massive , spiral galaxies at @xmath34 that may recently have harboured active agn , such as seyferts , but are currently in the less active or off phase .
we demonstrate that increases of at least a dex are possible for circumgalactic ( cgm ) @xmath5 time - averaged over fossil phases for normal seyfert luminosities turning on 1 myr out of every 10 myr .
our examples show just how sensitive metal lines can be to moderate agn , even if they are only on for a small fraction of time . as nearly all galaxies
are thought to contain supermassive black holes in their centres , fossil proximity zones may be very important for the interpretation of qso absorption line data .
this paper is organised as follows .
we introduce our non - equilibrium code and the methods used in this paper in [ sec : method ] .
we apply these methods to the agn proximity zones and proximity zone fossils for the examples mentioned above in [ sec : application ] , and we summarise in [ sec : summary ] . throughout
we use proper distances unless noted otherwise .
we use the non - equilibrium solver introduced in @xcite to follow the time - dependent ionisation states of 11 elements ( h , he , c , n , o , ne , mg , si , s , ca , and fe ) .
the solver also calculates ion - by - ion cooling and photo - heating to follow the time - dependent non - equilibrium net cooling , defined as cooling minus photo - heating .
the code is intended to be integrated in hydrodynamic codes , and includes a treatment to sub - cycle the integration of ionisation and cooling given the input courant timestep . in this work ,
we add the functionality to our method to allow the addition of photo - ionisation and photo - heating from a time variable source , which in this case has the spectral energy distribution ( sed ) of an agn .
the agn photo - ionisation rate ( s@xmath35 ) for an ionisation state @xmath36 of atomic species @xmath37 from a radiation source with specific intensity @xmath38 ( erg s@xmath35 @xmath39 hz@xmath35 ) is @xmath40 and the agn photo - heating rate ( erg s@xmath35 ) is @xmath41 where @xmath42 is frequency , @xmath43 is the ionisation frequency , @xmath44 is photo - ionisation cross - section , and @xmath45 is the planck constant . since we track 133 ionisation states of 11 elements , we integrate the above equations for the 122 ionisation states with electrons . for the egb , we replace @xmath38 with the ionising background sed averaged over the whole sky , @xmath46 , in equations [ eqn : gamma ] and [ eqn : eps ] , where @xmath47 has units of erg s@xmath35 @xmath39 hz@xmath35 sr@xmath35 .
as explained in @xcite , we calculate @xmath48 and @xmath49 at 50 different redshifts between @xmath50 using the ( * ? ? ?
* hereafter hm01 ) egb and the same @xmath44 used in cloudy ver . 10.00 last described in @xcite .
we use the hm01 egb as our fiducial background since this results in good agreement with the observed column density distribution of @xmath2 @xcite and metal line strengths and ratios @xcite . the time dependent evolution of the particle number density , @xmath51 , for each ionisation state , @xmath52 , is given by @xmath53 where @xmath54 is the free electron density ( @xmath55 ) , @xmath56 is the total recombination rate coefficient ( radiative plus di - electric , @xmath57 s@xmath35 ) , and @xmath58 is the collisional ionisation rate coefficient ( @xmath57 s@xmath35 ) .
the agn subscript indicates that the coefficients have been calculated using our agn template spectrum explained below .
charge transfer and auger ionisation are also included in our calculations as explained in @xcite , but were omitted from equation [ eqn : ionstate ] for simplicity .
charge transfer has negligible effects for the species we explore as this is more important for cases with significant @xmath2 neutral fractions , which are not the focus of this work .
including auger ionisation increases the ionisation by a small amount for the lithium - like ions we explore , but the differences are of the order of 0.1 - 0.2 dex at most , and usually much less @xcite .
the net cooling rate per unit volume ( @xmath59 , erg @xmath55 s@xmath35 ) , is the difference between cooling and photo - heating from the egb and the agn , @xmath60 the units of the cooling efficiency per ion , @xmath61 are erg @xmath62 s@xmath35 , where we divided the cooling rate per volume ( erg @xmath55 s@xmath35 ) for ion @xmath63 by @xmath64 to achieve density independence of collisional cooling , which allows us to tabulate these rates as a function of temperature only .
figure [ fig : sed ] compares our agn template spectrum , @xmath38 to the hm01 egb , @xmath65 at the three different redshifts ( @xmath66 , 0.87 , and 0.27 ) nearest the three redshifts we consider in this paper ( @xmath33 , 0.9 , and 0.25 ) .
the hm01 egb strength at the lyman limit at these three redshifts is from high to low-@xmath15 log[@xmath67 ( @xmath68)]@xmath69 , and @xmath70 .
& 0.25 ) we consider , along with our agn template spectrum , normalised to @xmath71 .
the euv power law of the agn is @xmath72 between 1300 and 354 ev based on observations by @xcite .
the x - ray portion of the spectrum is assumed to follow the cloudy agn template ( @xmath73 until 100 kev and then @xmath74 at higher energies ) .
the four coloured vertical lines at the bottom indicate the ionisation potentials needed to reach the corresponding ion listed . ] in cyan is the agn sed normalised to @xmath75 .
since this is a point source , we can convert its flux into @xmath67 equivalent units by dividing by @xmath76 steradians to get @xmath77 , which has the same units as the egb .
hence , @xmath78 is the intensity the egb would have if it were equally bright as the agn at the lyman limit .
the spectrum has an euv power law with slope @xmath72 between 1300 and 354 ev based on observations by @xcite . at higher energies , the spectrum follows @xmath73 until 100 kev and then @xmath74 at higher energies , which follows the template cloudy agn spectrum .
the most important slope for our consideration is the euv ( i.e. 10 - 354 ev ) slope , because these energies correspond to the ionisation potentials of the metal species we consider and which are marked in figure [ fig : sed ] .
however , the x - ray portion of the spectrum is also important for two reasons : i ) photo - heating rates of species like @xmath79 and @xmath80 depend on the shape of the x - ray spectrum , and ii ) we convert to observed luminosities of agn based on their 0.5 - 2.0 kev luminosities ( @xmath81 ) . to convert these units to agn luminosities ,
the above flux ( @xmath82 ) corresponds to @xmath83 agn at 100 proper kpc , which is the luminosity one may expect from a local seyfert .
the seyfert would outshine the @xmath34 hm01 egb at 912 by 45% at 100 kpc , but because of the harder agn spectrum at the @xmath5 photo - ionisation edge , 114 ev , the agn would outshine the hm01 spectrum by a factor of @xmath84 , assuming no absorption of the agn spectrum .
in general , for our template agn spectrum the conversion from agn soft x - ray luminosity to equivalent egb intensity units is @xmath85 where @xmath86 is the proper distance from the agn .
we consider three situations spanning a wide range of redshifts , from @xmath33 to @xmath87 , that are applicable to recent observations . in all cases
we assume the gas is being irradiated by the hm01 egb for the given redshift , and that it initially has the equilibrium temperature for which photo - heating balances cooling .
this temperature corresponds to @xmath88 k assuming solar abundances for the densities we consider , @xmath89 .
the main reasoning behind our parameter choices is that simulations indicate that a significant fraction of diffuse metals reside along a locus in density - temperature phase space where photo - heating balances cooling ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , and that many commonly observed metal - line species arise from gas at these densities and temperatures ( e.g. * ? ? ?
although solar abundances are used , we will argue that our theoretical metal columns scale nearly linearly with metallicity . in [ sec :
app1 ] we discuss our first application : an agn turning on at @xmath90 , which is motivated by observations of the very strong @xmath6 system observed by @xcite toward pg1206 + 459 , which appears to reside near a galaxy showing signs of agn activity .
we pay particular attention to this example , because of the large number of metal - line species observed in this system and also because @xmath6 is one of the species most affected by agn ionisation .
we vary several parameters including metallicity , density , agn strength , and agn lifetime . in [ sec : app2 ] , we consider the effects on @xmath8 , @xmath91 , and @xmath5 columns at @xmath33 , both when a quasar is on and after it turns off .
we estimate the fraction of the volume of the igm altered by fossil zones , and argue that a large fraction of observed metal lines may be affected .
lastly , in [ sec : seyfert ] , we explore short duty cycles of @xmath34 agn that significantly enhance @xmath5 columns , even while the agn is off
. this case could be applicable to normal looking spirals that were recently seyferts , where residual non - equilibrium effects of agn ionisation enhance metal columns . the general behaviour that we find can be summarised as follows .
the enhanced photo - ionisation from the agn rapidly raises the ionisation states of metals , thus allowing high - ionisation species to exist at higher densities than they would in equilibrium with the normal egb . after the agn turns off
, the long recombination times of metals , along with the multi - ionisation state nature of metals and their significant ionisation fractions ( i.e. @xmath92 for multiple @xmath36 ) , mean that metal columns often only return to equilibrium on a timescale that is longer than the agn - on phase .
thus , a whole new range of ionisation models that have not been considered are allowed . in our fiducial example
, we consider the effects of an enhanced ionisation in the proximity zone around an agn that is on for 20 myr at @xmath90 .
this case is directly applicable to observations by cos at a redshift where a large array of both low- and high - ionisation species can be observed .
this example is motivated by the extremely strong @xmath6 ( @xmath93 ) absorber observed by @xcite at @xmath12 along the pg1206 + 459 sight line with well - aligned lower ionisation species .
a foreground galaxy , labelled @xmath94 by that work , found at the same redshift at an impact parameter of 68 kpc , shows signs of being a post - starburst galaxy with a @xmath95-band absolute magnitude of @xmath96 and even agn signatures from [ @xmath97 emission that likely can not be explained by star formation alone .
figure [ fig : z09_tevol ] shows the time evolution of gas enriched to solar metallicity with @xmath98 , which corresponds to an overdensity of 77 at @xmath90 ( assuming @xmath99 and a hydrogen mass fraction of 0.75 ) .
this choice represents a typical density of metal - enriched cgm gas .
the gas is assumed to lie at its equilibrium temperature , @xmath100 k , when we mimic the instantaneous turn - on of the agn by adding our template agn spectrum to the hm01 egb at that redshift .
we assume an agn flux of @xmath101 , which has the @xmath2 ionisation equivalence of a uniform egb with @xmath102 integrated over @xmath103 steradians ; this corresponds to @xmath104 the egb value and a total ionisation parameter log[@xmath105@xmath106 , where @xmath107 is the ratio of hydrogen ionising photon density over @xmath108 .
using our template spectrum and integrating over the soft x - ray band ( 0.5 - 2.0 kev ) , we calculate that this is the flux at 100 kpc from a @xmath109 erg s@xmath35 agn , which appears to be a reasonable assumption given the observed impact parameter of 68 kpc for galaxy @xmath94 . in the top panel
we show the column densities of several observationally accessible ions by assuming an absorber length of 74 kpc , which is the predicted size of an @xmath2 absorber at this density and temperature according to equation 3 of @xcite .
all column densities scale linearly with the assumed absorber size . as the agn turns on at @xmath110 , the @xmath2 column density and ionisation fraction drop nearly instantaneously by a factor of @xmath111 as @xmath112 is only 1100 yr for this field .
the ionisation timescales for the metal species , shown in the middle panel of figure [ fig : z09_tevol ] during the agn - on phase ( thick dotted lines ) are longer , @xmath113 myr , but still less than the assumed agn lifetime of 20 myr . during the agn - on phase , the species @xmath5 , @xmath6 , and @xmath7 first rise to a peak and then fall as they are first ionised from lower ions to these states and then to higher ions , owing to the large increase in the radiation field . on the other hand , @xmath8 , which initially comprises 27% of carbon , decreases dramatically as carbon
is ionised all the way to @xmath114 in under 10 myr .
the column densities in the agn - on phase are applicable to proximate quasar absorption line ( qal ) systems observed near the background source quasar , or to intervening qal systems where the local ionising field is significantly enhanced by foreground agn near the sight line to the background source quasar .
absorption systems with log[@xmath105@xmath106 will have @xmath6 column densities that are @xmath115 stronger than for the hm01 field where log[@xmath105@xmath116 at @xmath98 , while @xmath117 will be reduced by a factor of @xmath118 , and @xmath119 will be reduced even more .
@xmath120 can either increase or decrease depending on the agn lifetime .
@xmath7 is the most enhanced photo - ionised species shown , increasing more than @xmath121 in strength to @xmath122 .
the temperature increases by @xmath123 to @xmath124 k due to photo - heating by the strongly increased ionising field ( bottom panel ) .
the temperature increase results primarily from the ejection of highly energetic electrons by x - ray photons leading to high - ionisation metal species ( e.g. @xmath125 , @xmath79 , and @xmath80 ) . for this solar - enriched case
, metal photo - heating dominates over that from @xmath126 , which has a lower ionisation potential than metal ions .
the photo - heating of @xmath2 does not contribute to the temperature increase , because in ionisation equilibrium the increase in the radiation field is exactly compensated by the corresponding decrease in the neutral fraction , and the latter is too small to result in significant non - equilibrium heating .
the agn proximity zone fossil phase begins after the agn turns off and the ionisation field switches back to the uniform hm01 egb .
the solid thin lines in the middle panel indicate the recombination timescales of the various species . although @xmath23 yr , @xmath117 reaches equilibrium in a much smaller timescale , @xmath127 yr , because the equilibrium neutral fraction it has to return to is only @xmath128 .
in contrast , fossil ionised bubbles around dead quasars can persist for @xmath129 during reionisation since the equilibrium state is neutral @xcite ; however , we are considering only cases where an egb ionises the igm to a high level .
@xmath117 remains slightly smaller than before @xmath110 , because photo - heating increased the temperature of the gas , thereby reducing the recombination rate .
the cooling timescale of @xmath130 gyr means that the gas will take of order this time to approach the original equilibrium temperature and @xmath2 fraction . even though the helium - like metal ion recombination timescales ( solid lines in the middle panel of figure [ fig : z09_tevol ] ) are @xmath131 shorter than for @xmath21 , their recombination lags are @xmath132 longer than for @xmath2 , because the equilibrium ion fractions are a significant fraction of unity .
a case in point is the timescale of @xmath133 myr that it takes @xmath8 columns to approach their previous equilibrium values .
a series of recombination timescales of species between @xmath114 and @xmath125 determine the total time it takes @xmath8 to return to equilibrium , since carbon is primarily in @xmath114 when the agn turns off .
however , the longest is @xmath134 myr , shown as the green solid line in the middle panel .
@xmath5 is significantly enhanced during @xmath135 myr after the agn turns off , even though its column density is decreased during the agn - on phase .
@xmath5 achieves at least twice as large columns between 33 - 74 myr , as oxygen recombines from a majority of @xmath80 at @xmath136 myr to @xmath9 at @xmath137 myr .
@xmath138 myr is the bottleneck slowing the recombination sequence ( orange solid line in the middle panel ) .
the most dramatic residual observable change is for @xmath6 , for which columns increase 100 fold for the first 15 myr of the fossil phase and remain at least @xmath139 stronger for 50 myr . despite the main recombination bottleneck being @xmath140 myr , or about half
as long as @xmath141 , @xmath6 takes several recombination times to approach its very low equilibrium value .
thus @xmath142 , which has been observed and attributed to collisionally ionised gas at @xmath143 k @xcite , could have an alternative colder origin if nearby galaxies currently or recently harboured an agn .
in passing , we note the recent work of @xcite , who find extremely strong high ionisation lines ( e.g. @xmath6 , @xmath7 ) in associated qso absorbers , which are possibly photo - ionised by the agn ; however , we note that those absorbers are probably located much nearer to the supermassive black hole than the extended proximity zones that are the focus of the present study . however , the same potential non - equilibrium fossil effects would apply , albeit on a shorter timescale given that the densities are likely much higher .
our method presented here is general enough to be applied to short - term qso fluctuations that keep these zones over - ionised relative to equilibrium solutions .
we have considered the very specific example of an agn proximity zone fossil applicable to enriched gas at densities typical of the circumgalactic medium .
while our choices appear reasonable , we wish to broaden the range of physical parameters and situations by varying the metallicity , density , agn strength , and agn lifetime .
we will relate these variations to situations that may occur in the real universe and will argue that fossil zones could significantly change the interpretation of metal qal observations .
we will continue to assume that the metals are cool ( @xmath144 k ) , residing at the equilibrium temperature where photo - heating balances radiative cooling .
* metallicity : * choosing solar metallicity for an absorption system is not unprecedented as detailed analysis of observed systems often yield solutions with solar or super - solar metallicities @xcite . however , many of these systems are attributed to lower density gas , which may on average be enriched to lower levels if metallicity increases with density as observations and simulations indicate ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . lowering the metallicity does not alter the recombination timescales or behaviour significantly . the equilibrium temperature in the hm01 ionisation field increases from @xmath100 k for @xmath145 to @xmath146 k for @xmath147 , owing to reduced metal - line cooling , and this results in only a 0.2 dex reduction of @xmath117 .
hence , metal - line strengths as a function of time can , to first order , be scaled with @xmath148 . however ,
if intergalactic metals are poorly mixed on small scales as some observations and simulations indicate @xcite , then high metallicities may even be appropriate at low densities .
* density : * changing the density alters the non - equilibrium behaviour because the recombination times scale as @xmath149 .
the ionisation parameter also changes when assuming a constant radiation field , meaning that in equilibrium atomic species are photo - ionised to higher ion states at lower densities .
we explore here four densities : @xmath150 , @xmath151 , @xmath152 , and @xmath153 corresponding to @xmath154 overdensities @xmath155 , 24 , 77 , and 243 respectively .
we assume an absorber length set by the @xcite relation ( @xmath156 kpc ) , which equals the jeans scale in a gas of uniform density and temperature .
we set @xmath157 k in the previous equation for simplicity , even though temperatures vary .
this size scale , which ranges from 230 to 42 kpc for our densities , has been shown to reproduce the observed sizes of the @xmath16 absorbers ( e.g * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and agrees very well with predictions from hydrodynamic simulations @xcite .
figure [ fig : z09_dens ] , top panel , shows that the @xmath117 of the absorbers initially range from @xmath158 with increasing density , which correspond to observed @xmath16 absorbers that typically exhibit detectable metal - line absorption @xcite .
as the agn radiation field turns on and off , @xmath117 reduces and increases by a factor @xmath118 on timescales that are too short to appear in the figure .
a slight deviation occurs for higher densities where photo - heating raises the temperature during the agn more , owing to the greater number of bound electrons available to ionise from metals .
although lower densities have a smaller photo - heating temperature increases ( bottom panel ) , the longer cooling times of several gyr mean that a 20 myr agn phase has longer lasting effects on the temperature of the igm .
using an agn with @xmath159 turning on at @xmath110 for 20 myr .
the assumed absorber lengths decrease from 164 to 29 kpc as density increases from @xmath150 to @xmath153 .
@xmath117 , @xmath119 , @xmath120 , @xmath160 , and temperature evolution are shown from top to bottom .
ionisation timescales remain unaltered between the four cases , but recombination times scale as @xmath149 meaning that the fossil effect will last longer at lower density .
metallicity is assumed to be solar in all cases , but metal columns scale with metallicity with the minor exception that photo - heating will be lower at lower metallicity and temperatures will not vary as much . ]
the third panel in figure [ fig : z09_dens ] shows @xmath120 for the four densities .
the overdensity range we explore covers nearly the same range that simulations predict observed @xmath5 absorbers to arise from at @xmath161 ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . when the agn turns on , the ionisation timescales are independent of @xmath108 , but the new ionisation equilibrium does depend on @xmath108 . for @xmath162 oxygen
is ionised from @xmath5 to higher states reducing these columns , while for @xmath163 lower ionisation oxygen is ionised to @xmath5 increasing the columns .
when the agn turns off , lower densities take longer to return to ionisation equilibrium because recombination times scale inversely with @xmath108 , thus , while the lowest density will have deficient @xmath5 relative to the equilibrium case for @xmath164 myr , the highest overdensity will have enhanced @xmath5 for a shorter timescale , similar to the agn lifetime of 20 myr in this case .
in contrast to @xmath5 , @xmath119 is reduced by the agn for all densities while @xmath160 shows an increase for @xmath165 ( 2nd & 4th panels of figure [ fig : z09_dens ] respectively ) .
the @xmath8 ionisation fraction peaks at @xmath166 when photo - ionised by the @xmath90 hm01 field at @xmath13 k. turning on the agn shifts this peak to @xmath167 .
hence , @xmath8 will be reduced by the agn , and the lowest densities will take the longest to recombine from higher ions in the fossil zone .
conversely , the ionisation fraction of @xmath6 peaks at @xmath168 at @xmath13 k , which shifts to @xmath169 with the agn .
hence , gas with @xmath170 can have @xmath160 columns as high as @xmath171 during the agn phase , 12,000@xmath19 larger than for the hm01 field in equilibrium .
the recombination times of a few myr means that this strong @xmath6 will be shorter lived in the fossil zone .
* agn radiation field increase : * in figure [ fig : z09_strength ] , we consider field increases , @xmath172 of factors 4.4 and 440 to simulate the agn proximity effect at 3.16 and @xmath173 the fiducial radius ( i.e. 316 and 32 kpc ) respectively for the same strength agn . or alternatively , an agn 0.1 and @xmath139 as strong at 100 kpc .
we again use our fiducial density @xmath174 .
the ionisation timescales decrease in proportion to the field increase , and stronger radiation fields cause metals to become more ionised in a shorter amount of time .
however , post - agn metal - line strengths show qualitatively similar evolution , because the recombination timescales are the same .
there is a greater delay for stronger fields , reflecting the additional recombination time from even higher ionisation states ( e.g. @xmath175 , @xmath176 ) . even for @xmath177 , which corresponds only to a factor @xmath178 increase relative to the egb ,
significant enhancements are seen , especially in @xmath6 .
the field increase is insufficient to ionise to @xmath179 , but post - agn @xmath6 columns are enhanced 100@xmath19 and decline slowly according to @xmath180 .
considering that the @xmath139 smaller field increase applies to @xmath181 more volume , the weaker field case could be most relevant to intervening qal , which provide a volume - weighted sampling of the igm . with @xmath98 and @xmath182 kpc for an agn turning on at @xmath110 for 20 myr .
the agn flux enhancement relative to the hm01 egb ranges from a factor of 4.4 ( long dashed lines ) to 440 ( short dashed lines ) .
the stronger the agn flux , the higher the ionisation state that the metals reach , and the more delayed the recombination to equilibrium becomes owing to the need to recombine through more ionisation states with significant recombination times .
although the metallicity is assumed to be solar , the metal columns scale with metallicity , except for the minor effect of photo - heating being less efficient at lower metallicity . ]
* agn lifetime : * thus far we have assumed a single @xmath183 myr lifetime , but agn lifetimes are predicted to be as short as @xmath184 yr .
figure [ fig : z09_life ] shows what happens if an agn is on for shorter periods at multiple times .
we choose to show an agn that is on 5 out of every 50 myr yielding a duty cycle , @xmath185 , and an agn that is on 1 out of every 25 myr , yielding @xmath186 . for metals ,
the effect of a shorter lifetime is similar as that of a weaker agn : they are not as highly ionised by the agn , but the post - agn recombination times are the same . a shorter @xmath4 results in metals not reaching as highly ionised states .
post - agn , the metals are less likely to have to recombine through these upper states reducing the number of recombination times to return to equilibrium . nevertheless , the most significant bottleneck is the longest metal @xmath20 from the helium - like state ( @xmath125 , @xmath79 , @xmath179 ) to the lithium - like ions we plot .
hence metal - line evolution is often similar in these three cases although shifted in time depending on the strength and of course the frequency of agn - on phases . with @xmath187 and @xmath188 kpc for an agn flux of @xmath189 .
we compare our fiducial case of a single agn - on phase of 20 myr with an agn - on phase of 5 myr every 50 myr and 1 myr every 25 myr .
the first agn on phase always starts at @xmath110 .
significant non - equilibrium behaviour occurs in all cases significantly altering the time - averaged column densities during the agn - off phase .
photo - heating in bursts can boost average gas temperatures as well .
although the metallicity is assumed to be solar , the metal columns scale with metallicity , except for the minor effect of photo - heating being less efficient at lower metallicity . ]
even short agn lifetimes can efficiently enhance metal columns if the duty cycle is similar to or smaller than the recombination times .
a case in point is @xmath6 for @xmath190 myr in the 4th panel , whose column density is enhanced by @xmath191 averaged over 0 - 200 myr , despite the agn being on only 4% of the time . for @xmath192 myr ,
@xmath5 is enhanced by a factor of three and remains relatively constant , owing to a coincidence : @xmath193 myr is similar to @xmath4 , and @xmath194 myr is similar to the duty cycle timescale of 25 myr .
the implications for small duty cycles with resonances that amplify the fossil effect will be explored further in
[ sec : seyfert ] . finally , the bottom panel shows how duty cycles can pump temperatures up , owing to the combination of short photo - heating timescales and long cooling timescales if the gas has solar metallicity .
multiple agn - on phases are applicable to the real universe , although it is unlikely that an agn will have such rhythmic repetitions as our examples here . instead , an agn phase may last on the order of @xmath195 yrs , but fluctuate significantly .
we suggest a model for galaxy @xmath94 in [ sec : pg1206 ] where it was much brighter in the recent past , perhaps even a bright qso , while today it still is a much weaker agn .
the point of the duty cycle examples is to show the fossil effect is non - linear , where multiple short agn phases have a much greater effect than one long agn phase .
* extra - galactic background and agn euv power laws : * we have chosen an agn spectrum with an euv power law of @xmath72 based on @xcite . the local agn ionises diffuse gas that is additionally ionised by the hm01 spectrum , which itself assumes a qso euv input of @xmath196 . when we apply either the @xcite or ( * ? ? ?
* hereafter hm12 ) egbs that use hard qso input spectra of @xmath197 and @xmath72 respectively , then the increase in the radiation field at the ionisation potentials of high ions like @xmath5 , @xmath6 , and @xmath7 is more muted , but the timescales for recombination are essentially the same . our fiducial example in figure [ fig : z09_tevol ] has the following columns in equilibrium with the hm12 egb : @xmath198 ( 0.2 dex increase relative to hm01 ) , @xmath199 ( 0.4 dex decrease ) , @xmath200 ( 0.6 dex increase ) , @xmath201 ( 0.8 dex increase ) , and @xmath202 ( 0.4 dex increase ) .
the addition of the agn field with @xmath102 leads to the same agn - on phase columns as in the fiducial example , meaning that relative to equilibrium with the hm12 egb , @xmath2 now declines by 100-fold ; the @xmath8 decrease , @xmath6 increase , and @xmath7 increase are all less ; and finally @xmath5 decreases more .
the metal lines re - equilibrate during the fossil phase on the same recombination timescales , but @xmath6 is only @xmath203 higher than the hm12 egb equilibrium in the first 15 myr after the agn turns off , instead of 100-fold increase relative to the hm01 field .
the observational consequences are the same , except that the egb equilibrium columns for @xmath6 are more likely to be observable , and fewer recombination timescales are required to reach egb equilibrium .
hence , the biggest difference in using hm12 or any other background is that photo - ionised equilibrium columns are different . in the real universe
, there exists an extremely large range of observed agn euv power laws . @xcite and
@xcite together find a range of @xmath204 to @xmath205 encompassing most of their measured agn , and even an outlier with @xmath206 , which is the well - known quasar he2347 - 4342 at @xmath207 that has been used as a probe of @xmath126 reionisation @xcite .
hence , the fossil effect may be even greater for harder agn , which are more capable of ionising high ions , while softer agn will have much less of an effect .
interestingly , @xcite provide a sample showing an anti - correlation between euv hardness and agn luminosity .
their composite spectrum has a harder euv slope , @xmath208 , than @xcite and @xcite , which suggests greater fossil effects .
in fact , they find that their lowest luminosity agn have euv slopes exceeding 0.0 in the low - redshift universe , which makes our exploration of weak agn in [ sec : seyfert ] possibly even more relevant .
hence , our choice of a harder agn euv spectrum relative to the egb for our fiducial example may be quite applicable and could even underestimate the effect .
we apply our model to the @xmath12 pg1206 + 459 absorber in figure [ fig : z09_ratio ] , where we plot the evolution of the column densities of metal - line species and @xmath2 observed by @xcite .
we explore a range of densities , agn flux increases , and metallicities , finding that the model with @xmath209 , @xmath210 , and @xmath145 provides a surprisingly good fit for the ions observed by @xcite using a single gas phase .
we plot the evolution of one ion against another ion starting at @xmath110 when the agn turns on ( large magenta diamond ) , through the 20 myr agn - on phase ( magenta ) , and the fossil phase ( beginning with a large blue diamond at 20 myr and then changing colours over the next 50 myr according to the colour bar ) .
diamonds are plotted every myr .
the open black squares show the @xcite component absorbers , indicating lower limits for saturated absorbers as triangles in the case of @xmath9 and @xmath211 , and upper limits as upside - down triangles for non - detections of @xmath7 .
we combine their @xmath212 and @xmath213 components into o motivated by the blended @xmath6 detection .
the top left panel shows that we can reproduce the observed @xmath6 and @xmath10 columns as arising from the same phase 4 - 5 myr after the agn turns off , which is how we constrained our parameter choice .
no equilibrium solution with the hm01 background can reproduce both @xmath6 and @xmath10 , which are shown by @xcite to arise from very different densities and temperatures in simulations of the cgm that assume ionisation equilibrium .
the agn - on phase makes @xmath6 columns large enough , but @xmath10 is orders of magnitude too weak .
the fossil phase allows a solution for both species that persists in this case for @xmath214 yrs given the spacing of the diamonds .
considering other ions in figure [ fig : z09_ratio ] , our solution agrees with the @xmath7 upper limits ( top middle panel ) , and because @xmath2 re - equilibrates to the hm01 solution very rapidly we are able to reproduce the high neutral hydrogen columns .
neither of these ions are reproduced during the agn - on phase .
several other observed ions are plotted against @xmath6 in the second row : @xmath91 is somewhat high , but could be rectified by not assuming solar abundances as we do here , @xmath9 lower limits agree , and @xmath215 does appear to be over - abundant by a dex .
the bottom six panels show mostly lower ions , either plotted against @xmath10 or two ions of the same atomic species plotted against each other .
the agreement with @xcite is usually quite good at @xmath216 myr into the fossil phase , although nitrogen and silicon may have lower than solar abundances , and sulphur appears to be greatly over - abundant , by as much as a dex .
the only species not shown from @xcite is @xmath217 , for which our model reproduces the lower limits .
we plot @xmath5 and @xmath8 as well ( bottom right panel ) , even though these species have not been observed at high resolution , making column density determinations not yet possible ; however @xcite observations indicate very saturated profiles , which agree with our predictions that @xmath218 and @xmath219 .
figure [ fig : z09_ratio ] shows merely one possible time - dependent history for an agn ionising cool , metal - enriched gas .
our fiducial example in figure [ fig : z09_tevol ] can also fit @xmath6 and @xmath91 using @xmath220 gas between 20 - 50 myr after the agn turns off , but requires a separate phase for the low ions as @xmath10 is 1000@xmath19 too weak .
figure [ fig : z09_ratio ] shows a model with a 140-fold increase in flux at the lyman limit , but models with 0.5 dex weaker and stronger agn show nearly identical tracks .
a proximity zone fossil can enhance metal lines over a significant path length if the agn was strong in the past .
we showed that a wide range of agn intensities , as would be expected along a large path length through a proximity zone , will lead to similar recombination time histories and hence similarly enhanced columns for high - ionisation species at a given density .
if a proximity zone is many mpc in size , then enhanced metal columns should be observed along a similar path length .
a fossil zone could be traced by unusually strong high - ionisation species over a large velocity range as in the case of the @xmath6 in the @xcite absorber , which extends over 1500 @xmath17 .
this corresponds to a hubble flow of over 10 proper mpc implying an extended fossil zone and a very strong agn in the past .
we choose an agn lifetime of 20 myr , but given that equilibrium agn - on columns are achieved in only 1 myr ( i.e. most of the agn - on evolution in figure [ fig : z09_ratio ] occurs before the first small magenta diamond appears ) , the fossil phase evolution will be similar for a shorter lived agn .
our point is that models with time - dependent agn that have either turned off in the past , recently turned on , or fluctuate are worth considering , especially for an extreme case such as the pg1206 + 459 absorber with evidence of a post - starburst agn within 100 kpc of the sight line . in summary
, we propose a new time - dependent model for the @xmath12 pg1206 + 459 absorber that relies on enhanced photo - ionisation of @xmath221 k gas by an agn that turned off several myr in the past .
unlike equilibrium solutions , out model can reproduce the columns of @xmath6 , @xmath10 , @xmath2 , and other metals using a single phase solution that lasts @xmath222 myr .
this scenario provides an alternative to the outflow shell model proposed by @xcite of a @xmath223 k hot gas component and a velocity - coincident cold gas component , which may be expected for a propagating shock with post - shock radiatively cooling metal - enriched gas . in reality , gas at a variety of densities , and hence decay times , may contribute to the absorber .
however , we note that the choice of @xmath224 is well motivated by ionisation solutions to @xmath13 k clouds observed by @xcite at @xmath225 .
our model requires @xmath226 of gas per component , assuming a path length through the centre of a spherical @xmath227 kpc , @xmath228 solar - enriched cloud .
this is @xmath229 less mass per component than the @xcite thin shell model for their @xmath223 k phase .
another application of our code is the transverse proximity effect ( tpe ) where a foreground qso ionises gas along the sight line toward a background qso .
recent work has been performed by @xcite , @xcite , and @xcite who explored the tpe at @xmath230 , near the peak of qso activity .
the tpe can be elusive when traced by @xmath2 alone owing to complicating effects of the overdense regions creating more @xmath2 absorption , which goes in the opposite direction of decreased @xmath2 owing to the qso ionisation ( see also * ? ? ?
for example , @xcite argue that qso emission is anisotropic based on the enhancement of lyman - limit systems in transverse sight lines indicating that the qso radiation does not effect off - sight line directions as often , even after accounting for the fact that qsos may live in overdense regions .
@xcite analysed metal - line systems in a quasar triplet at @xmath231 and found that the tpe was apparent when @xmath8 and @xmath5 aligned with @xmath2 are included in the analysis .
they in fact argued that unusual @xmath232 and @xmath233 ratios of components in transverse proximity zones indicate that the qso radiation field reaches off - sight line directions , and that the two foreground qsos , which are extremely bright and rare , more likely radiate isotropically .
their work shows that the non - linear ionisation properties of metal lines , when used in conjunction with @xmath2 , can help constrain the tpe ( see also * ? ? ?
* ; * ? ? ?
we turn on our agn template spectrum for 20 myr and find the average change in column density owing to the enhanced field . the left tables in figure [ fig : z25_qso_after ] show , from top to bottom , the average enhancement in @xmath2 , @xmath8 , @xmath91 , and @xmath5 of the time - averaged column densities during the 20 myr qso - on phase for a grid of densities ( log[@xmath234 ) and fluxes ( log[@xmath235 ) . the second column in each table shows equilibrium column densities ( log[@xmath236 ) using the hm01 @xmath33 egb , absorber lengths according to @xcite , and solar metallicity . as with all cases in this paper , we assume the gas initially rests at the temperatures where photo - heating balances cooling for the hm01 field , which fall from @xmath237 k as density increases from @xmath238 at @xmath33 .
the grid cells to the right of the hm01 column indicate the column density change @xmath239log[@xmath240 ( in dex ) relative to the hm01 equilibrium , colour - coded to be redder if the qso decreases the column and bluer if it is increased .
the behaviour of @xmath2 is relatively easy to understand . as the agn turns on
, @xmath117 equilibrates to a column density that is inversely proportional to the intensity of the new radiation field .
we define the agn enhancement factor as @xmath241 with the @xmath33 hm01 egb having log[@xmath67]=@xmath243 combined with the weakest agn field log[@xmath172]=@xmath245 , @xmath2 is reduced by log[@xmath246@xmath247 dex .
the six agn field enhancements we list range from @xmath248=1.6 - 200 , which decrease @xmath117 in the same proportion modulo the effect of photo - heating raising the temperature of the gas and thus further reducing @xmath2 .
photo - heating by the strongest agn reduces @xmath2 by an additional factor of 2 - 3 in @xmath249 gas , owing to the ejection of energetic metal electrons at x - ray energies .
the behaviour of the metal lines is not as simple as that of @xmath2 .
as we saw in the previous subsection , agn almost always ionise @xmath8 to higher states and thus reduce their columns , while enhancing @xmath5 at higher @xmath108 and reducing it at lower @xmath108 , depending on the field strength .
we again assumed solar metallicity , but the listed @xmath239log[@xmath240 values are applicable to other metallicities modulo the effect of increased photo - heating at higher metallicity .
the same trends are observed by @xcite : increased @xmath5 and decreased @xmath8 .
their @xmath248 ranged from @xmath250 and they derived densities of @xmath251 , which would correspond to the bottom four rows and right - most four columns in the left tables of figure [ fig : z25_qso_after ] .
one difference is that they use a single power law of @xmath252 in their uv egb , while we are using the hm01 field enhanced with a @xmath72 agn spectrum plus an x - ray component .
this allows our field to ionise @xmath5 to @xmath79 and above , which explains why we find more reduction in @xmath120 for the stronger fields : @xmath5 is ionised to higher states much more easily .
an additional point about the tpe during the agn - on phase is that @xmath5 increases significantly more than @xmath2 decreases at higher densities and lower field strengths ( see blue cells in bottom left panel of figure [ fig : z25_qso_after ] ) .
we may thus expect significantly enhanced @xmath5 with weaker than usual @xmath2 toward the edge of a proximity zone , which may approach @xmath253 proper mpc in size for the brightest qsos .
in fact , the large enhancement in proximate @xmath5 absorbers observed by @xcite within @xmath254 @xmath17 of the brightest qsos at lower redshifts ( or @xmath255 mpc at @xmath256 ) , combined with their lower observed @xmath257 ratios agrees with this prediction .
we put forth a model that these are not absorbers directly associated with the qsos , but instead intervening igm absorbers on the outskirts of the proximity zone with a @xmath248 of a few an perhaps even less than two .
this of course assumes that the agn has a harder euv spectrum than the egb .
a similar point can be made that @xmath91 could be enhanced in agn proximity zones . only cells with @xmath258 and weaker agn fields
show noticeably enhanced @xmath91 when the agn is on , but these cells may be well - represented in qso proximity zones . consider that while @xmath259 for the equilibrium hm01 field peaks at @xmath170 ( assuming @xmath260 as in figure [ fig : z25_qso_after ] ) , the highest observed @xmath91 column may correspond to @xmath261 given that metallicity likely correlates with density .
factoring in that most of the proximate volume and path length corresponds to the weaker fields , the strongest observable @xmath91 columns could correspond to the outskirts of proximity zones .
since nitrogen is less abundant than other metals and @xmath91 is much rarer than @xmath8 and @xmath5 , this agn enhancement could boost normally unobservable or hard - to - detect @xmath91 columns by @xmath262 into obvious detections .
we now cross over to the fossil phase by considering the time - averaged columns of @xmath2 and metals during the first 20 myr after the qso turns off .
this timescale is chosen to demonstrate that significant fossil effects can persist on a similar timescale as @xmath4 .
we show the deviations in columns on the right side of figure [ fig : z25_qso_after ] .
listed is the time - averaged column deviation over this time span ( note that the actual deviation will likely be larger or smaller at a specific fossil zone time ) .
the @xmath8 column is almost always reduced , but not to the same extent as during the qso - on phase .
nonetheless , if a significant fraction of intervening @xmath8 absorbers reside in fossil zones , then models that try to fit these assuming a uniform egb like hm01 ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) would underestimate the amount of carbon .
hence , the true global ionisation correction of @xmath8 would be larger , and the inferred igm metallicity would be too low
. the opposite would be true for @xmath5 at higher densities , but it would have the same trend as @xmath8 if it mostly arises from lower density gas .
meanwhile , @xmath2 would remain nearly unaffected except for the slightly reduced columns owing to photo - heating of metals .
the resulting absorption line systems will have ratios ( e.g. @xmath263 and @xmath264 ) that would differ significantly from those predicted by equilibrium models with a normal egb or in a proximity zone .
we now present an estimate of the fraction of the volume of the universe that may be affected by transverse agn proximity and fossil zones at high-@xmath15 .
we estimate the luminosity function ( @xmath265 , # /mpc@xmath266/log(@xmath267 ) ) of agn at 912 using the analytic fitting formulae of @xcite at @xmath33 . for a given isolated agn ,
the proper radius within which the agn flux is at least @xmath268 times @xmath47 is given by @xmath269 , where @xmath267 is the agn luminosity at @xmath42 in ergs s@xmath35 hz@xmath35 .
hence , the fractional volume , @xmath270 , exposed to at least @xmath268 is @xmath271 when integrating over the luminosity function ( @xmath42 is in the rest frame ) .
the @xmath272 term is a result of @xmath273 being in comoving mpc and our proximity zone units being proper units . using @xmath274 at @xmath33 ,
the volume ionised by agn to at least the level of the hm01 background , @xmath275 , is 0.16% , and the volume ionised to @xmath276 is 0.005% .
this calculation does not depend on how anisotropically agn emit , because the average subtended solid angle over which agn of a given luminosity emit is compensated by the number of all emitting agn for that luminosity .
the @xmath270 fractions are much lower at lower redshift , where agn are less frequent and the universe has expanded . while the fraction of volume with @xmath277 is small compared to the volume filling factors of metals in simulations calibrated to fit observed @xmath8 statistics , which is of order 10% at @xmath278 @xcite , there are reasons to argue that agn proximity and fossil zones could affect a significant fraction of observed metal lines .
first , strong @xmath8 systems ( @xmath279 ) are associated with lyman - break galaxies ( lbgs ) at small radii ( 80 - 85 proper kpc * ? ? ?
* ; * ? ? ?
* ) , and the co - moving line density of such strong absorbers is consistent with being within 50 kpc of @xmath280 lbgs @xcite .
this implies that many of the strongest metal - line absorbers occupy the small fraction of volume with enhanced ionisation .
more importantly , the multiplicative effects of agn with short duty cycles can significantly expand the volume affected by fossil zones .
our lower redshift example in figure [ fig : z09_life ] shows that a @xmath186 duty cycle leads to 3@xmath19 greater @xmath5 and 5@xmath19 reduced @xmath8 columns time - averaged once the agn first turns on at @xmath110 .
this duty cycle is relevant since the observed fraction of lbgs with active agn is @xmath281 @xcite , which could reflect their duty cycle .
taking @xmath282 for @xmath186 means that the fossil effect applies to a @xmath283 greater volume than the proximity zone effect , and hence that 4% of the total volume is affected by @xmath284 .
this volume fraction is similar to that thought to be required to account for the observed incidence of weak metal lines @xcite suggesting that non - equilibrium agn fossil zones may be critically important .
finally , metal lines can be enhanced if @xmath285 , as is shown in figure [ fig : z25_qso_after ] for log[@xmath172]=@xmath245 , or @xmath286 .
this leads to a doubling of the volume affected by both proximity and fossil zones , and @xmath8 is reduced by @xmath287 in almost every case .
again , this arises because the agn euv spectrum is harder than that of the egb .
we conclude that a large fraction of metal lines could arise from fossil zones at high-@xmath15 , but precisely what fraction depends on many factors .
if the agn - on phase is long - lived and singular for most agn then the fossil effect would be limited . however
, if agn are fluctuating with short duty cycles in typical star - forming galaxy hosts and have hard euv slopes , then the fossil effect could be a critical consideration for typical metal lines . we may have to re - evaluate the physical interpretation of @xmath8 and @xmath5 absorbers at high-@xmath15 , and the amount of underlying metals traced in diffuse gas . [
sec : app3 ] our last case study are low - redshift metal absorbers associated with weaker agn , such as seyferts .
we consider an admittedly idealised situation of a 10% duty cycle where the agn turns on for 1 myr out of every 10 myr at @xmath34 .
we plot the time history of @xmath5 in figure [ fig : z025_seyferthist ] for three different field enhancements , to show how the continuous pumping resulting from a short duty cycle can even significantly enhance the ionisation around weak agns .
we again assume @xmath98 solar - enriched gas at the equilibrium temperature of @xmath288 k. columns ( top panel ) and gas temperature ( bottom panel ) for a @xmath98 , @xmath289 kpc solar metallicity absorber at @xmath34 , where an agn of the listed intensity turns on for 1 myr every 10 myr starting at @xmath110 .
the periodic agn activity significantly enhances @xmath5 columns and doubles the gas temperature over 200 myr in the most extreme case via photo - heating of metals .
this case could be applicable for metals around normal looking galaxies if agn duty cycles are short and have frequencies comparable to the @xmath5 recombination timescale . ]
this situation may be applicable to markarian galaxies in the local universe .
these are among the biggest , brightest spirals , and have an unusually blue central core that usually requires non - thermal emission , which often indicates agn activity .
about 10% of markarians are seyferts @xcite , which motivates our choice of a 10% duty cycle . of course , not all markarians are seyferts with a short duty cycle , but we consider the possibility that @xmath29010% of markarians were more active agns in the recent past ( i.e. in the last @xmath253 myr ) .
the three cases plotted in figure [ fig : z025_seyferthist ] , @xmath291 , @xmath292 , and @xmath293 correspond to soft x - ray luminosities at 100 proper kpc of @xmath294 , @xmath295 , and @xmath296 agn , respectively , using our agn template spectrum .
these luminosities are in the range of seyfert x - ray luminosities observed by @xcite .
the bottom panel of figure [ fig : z025_seyferthist ] shows that the temperature doubles to @xmath297 k in the strongest case .
we tabulate the time - averaged enhancements of @xmath8 , @xmath5 , and @xmath6 over 200 myr , or the first 20 cycles , for a grid of densities ( log[@xmath108/@xmath298 ) and agn fluxes ( log[@xmath299/(@xmath68 ) ] ) in figure [ fig : z025_seyferttable ] .
similar to figure [ fig : z25_qso_after ] , the second column in each table shows column densities ( log[@xmath300/@xmath301 ) using the hm01 egb ( here at @xmath34 ) , absorber lengths according to @xcite , @xmath302 k before the agn turns on , and solar metallicity .
the grid cells to the right indicate the column density change @xmath239log[@xmath240 relative to the hm01 equilibrium , colour - coded to be redder if the column density goes down due to the agn and bluer if it goes up .
the @xmath108 densities correspond to overdensities ranging from 27 to 8500 .
while we show solar metallicity , the change is similar for lower metallicities .
the main point of this diagram is to show that even for a 10% agn duty cycle , the time - averaged columns are significantly changed , mostly in a positive direction for these higher metal ions . while there has never been a survey designed to probe the halos of galaxies showing signs of on - going or recent agn activity in the low - redshift universe , it may be worth exploring in light of the cos observations of @xcite revealing very strong @xmath5 around normal spiral , star - forming galaxies at @xmath303 .
markarian galaxies , for example , may show even stronger @xmath5 if this gas originates from a photo - ionised component at @xmath304 k.
the assumption of a uniform extra - galactic background ( egb ) breaks down in an agn proximity zone , which is over - ionised relative to the rest of the igm .
@xmath2 columns are reduced and metals are ionised to higher states owing to increased photo - ionisation .
we demonstrate here that once the agn turns off , metals residing in `` fossil '' agn proximity zones may remain over - ionised for timescales exceeding the typical agn - on phase . using the non - equilibrium code introduced in @xcite , we follow metal - enriched gases initially ionised by the egb @xcite under the effect of increased ionisation from a local agn . after the agn turns off
, metal ions in fossil proximity zones often require many myr to return to the equilibrium values , owing to the significant recombination times of metal ion species .
we demonstrate how this non - equilibrium fossil effect alters observational metal - line diagnostics used to constrain the physical state of intergalactic and circum - galactic gas .
in particular , we show that commonly observed lithium - like ions ( i.e. @xmath8 , @xmath91 , @xmath5 , @xmath6 ) can be heavily affected at a range of redshifts from the present - day to z@xmath305 .
this behaviour is in contrast to @xmath2 , which re - equilibrates comparatively rapidly once an agn turns off , owing to its equilibrium state having very low neutral fractions in the igm .
we consider metal - enriched gas with densities @xmath306 at @xmath307 and 0.25 , which corresponds to diffuse igm and cgm gas , that is initially in thermal equilibrium ( @xmath30 k ) .
our parameter choices are motivated by hydrodynamic simulations , which show a significant fraction of diffuse metals in thermal equilibrium , where cooling balances photo - heating , owing to short cooling times from higher temperatures ( @xmath308 k ) .
the gas begins in ionisation equilibrium with the hm01 egb , and becomes irradiated by an agn template spectrum turned on for 1 to 20 myr corresponding to an ionisation increase of between @xmath309 the egb radiation field strength at the lyman limit .
our fiducial case corresponds to a local agn turning on at @xmath90 , a redshift for which a wide range of metal lines are observable with current facilities ( e.g. cos on hst ) and which corresponds to the @xcite observations of very strong @xmath6 absorption in a system associated with a galaxy exhibiting post - starburst and possible agn signatures .
we show that for reasonable agn energies , @xmath6 columns are increased by a factor of @xmath310 during a 20 myr agn - on phase and also during the 15 myr in the following fossil phase , and remain @xmath311 stronger for 50 myr after the agn turns off .
we show for a variety of gas densities , agn strengths , and agn lifetimes that @xmath6 columns of @xmath312 are possible for significant timescales during the fossil phase for solar abundances .
we directly model the components of the @xmath12 pg1206 + 459 absorber @xcite , finding a single phase model that successfully fits @xmath6 , @xmath10 , @xmath2 , and other metal columns assuming @xmath313 , @xmath13 k gas in a fossil zone several myr after a strong qso turned off .
our application to this absorption system demonstrates that a whole new class of non - equilibrium solutions are available for metal - enriched diffuse gas in the presence of evolving ionisation fields .
we also model agn ionisation at @xmath231 , including the transverse proximity effect for a foreground qso zone in a qso pair , where observed @xmath264 and @xmath263 line ratios indicate enhanced ionisation over the normal egb @xcite .
we first show that active proximity zones can exhibit these unique ratios , and then consider how the non - equilibrium fossil effect can alter @xmath8 and @xmath5 columns for timescales longer than the agn lifetime .
unique line ratios are possible where @xmath2 has rapidly re - equilibrated but the metal lines are still recombining .
we argue that it is plausible that fossil zones could affect many more absorbers in the high-@xmath15 universe than agn proximity zones if agns fluctuate with short duty cycles , have hard extreme uv slopes , and are associated with typical star - forming galaxies , which prolifically enrich the igm .
a large fraction of intervening metal absorbers in typical qso spectra may well be affected by the fossil effect , meaning that ionisation corrections assuming a uniform egb would be wrong .
@xmath8 columns are almost always reduced in fossils meaning that the true diffuse gas metallicity becomes higher than when one assumes equilibrium with the egb field .
finally , we consider lower redshift agn ionising their surrounding cgm with a 10% duty cycle where the agn is on for 1 out of every 10 myr . despite this short duty cycle ,
high ions ( @xmath5 and @xmath6 ) are significantly enhanced , often by more than 1 dex , and temperatures are increased by a factor of two for stronger fields .
the key to these increases of metal columns and temperature are that the photo - ionisation / photo - heating timescales are short or comparable to the agn lifetime , while the recombination / cooling timescales are long or comparable to the duty cycle interval .
the result is intriguing in that surprisingly little energy input is necessary to greatly enhance metal columns in the halo of fluctuating typical agn .
these non - equilibrium fossil and proximity zone effects may complicate the analysis of intergalactic metal lines that assume photo - ionisation equilibrium with a uniform egb .
for example , cosmological hydrodynamic simulations using such assumptions underestimate the observed number density of the highest equivalent width @xmath6 @xcite and @xmath5 absorbers @xcite , unless a sub - grid turbulence model is invoked post - run .
proximity zone fossils could provide an alternative model for such strong , highly ionised metal absorbers .
the fossil effect is unique in that @xmath2 re - equilibrates rapidly while metal lines remain over - ionised leading to different ionisation corrections to obtain metallicities
. this could be particularly important during the quasar era , @xmath314 , for which we find that the fraction of the volume occupied by fossil zones may be similar to the fraction that is thought to be enriched with heavy elements .
we are grateful for discussions with sebastiano cantalupo , arlin crotts , romeel dav , kristian finlator , martin haehnelt , zoltan haiman , phil hopkins , mike shull , chuck steidel , john stocke , todd tripp , and lisa winter .
we thank ali rahmati for a thorough reading of this manuscript , and the anonymous referee for a constructive review .
this work benefited from financial support from the netherlands organisation for scientific research ( nwo ) through veni and vidi grants , from nova , from the european research council under the european union s seventh framework programme ( fp7/2007 - 2013 ) / erc grant agreement 278594-gasaroundgalaxies and from the marie curie training network cosmocomp ( pitn - ga-2009 - 238356 ) .
we are also thankful for the hospitality provided by the university of colorado , boulder where part of this work was completed .
adelberger k. l. , shapley a. e. , steidel c. c. , pettini m. , erb d. k. , & reddy n. a. 2005 , apj , 629 , 636 aguirre , a. , dow - hygelund , c. , schaye , j. , & theuns , t. 2008 , apj , 689 , 851 altay , g. , theuns , t. , schaye , j. , crighton , n. h. m. , & dalla vecchia , c. 2011 , apjl , 737 , l37 bajtlik , s. , duncan , r. c. , & ostriker , j. p. 1988
, apj , 327 , 570 bolton , j. s. , & haehnelt , m. g. 2007 , mnras , 374 , 493 booth , c. m. , schaye , j. , delgado , j. d. , & dalla vecchia , c. 2012 , mnras , 420 , 1053 carswell , r. f. , webb , j. k. , baldwin , j. a. , & atwood , b. 1987 , apj , 319 , 709 cen , r. , & chisari , n. e. 2011 , apj , 731 , 11 cooksey , k. l. , kao , m. m. , simcoe , r. a. , omeara , j. m. , & prochaska , j. x. 2012 , arxiv:1204.2827 dav , r. , oppenheimer , b. d. , katz , n. , kollmeier , j. a. , & weinberg , d. h. 2010 , mnras , 408 , 2051 ding , j. , charlton , j. c. , churchill , c. w. , & palma , c. 2003 , apj , 590 , 746 faucher - gigure , c .- a . , lidz , a. , zaldarriaga , m. , & hernquist , l. 2009 , apj , 703 , 1416 ferland , g. j. , korista , k. t. , verner , d. a. , ferguson , j. w. , kingdon , j. b. , & verner , e. m. 1998 , pasp , 110 , 761 ford , a. b. , oppenheimer , b. d. , dav , r. , et al .
2013 , mnras , 432 , 89 furlanetto , s. r. , haiman , z. , & oh , s. p. 2008 , apj , 686 , 25 gonalves , t. s. , steidel , c. c. , & pettini , m. 2008 , apj , 676 , 816 haardt , f. & madau , p. 1996 , apj , 461 , 20 haardt , f. & madau , p. 2001 , in clusters of galaxies and the high redshift universe observed in x - rays , recent results of xmm - newton and chandra " , xxxvith rencontres de moriond , eds .
neumann & j.t.t .
haardt , f. , & madau , p. 2012 , apj , 746 , 125 haiman , z. , & hui , l. 2001 , apj , 547 , 27 hennawi , j. f. , prochaska , j. x. , burles , s. , et al .
2006 , apj , 651 , 61 hennawi , j. f. , & prochaska , j. x. 2007 , apj , 655 , 735 hopkins , p. f. , hernquist , l. , cox , t. j. , et al .
2006 , apjs , 163 , 1 hopkins , p. f. , richards , g. t. , & hernquist , l. 2007 , apj , 654 , 731 jakobsen , p. , jansen , r. a. , wagner , s. , & reimers , d. 2003 , a&a , 397 , 891 lehner , n. , savage , b. d. , richter , p. , sembach , k. r. , tripp , t. m. , & wakker , b. p. 2007 , apj , 658 , 680 martini , p. , & weinberg , d. h. 2001 , apj , 547 , 12 meiring , j. d. , tripp , t. m. , werk , j. k. , et al .
2013 , apj , 767 , 49 muzahid , s. , srianand , r. , arav , n. , savage , b. d , & narayanan , a. 2013 , arxiv:1302.5510 narayanan , a. , wakker , b. p. , & savage , b. d. 2009 , apj , 703 , 74 narayanan , a. , et al .
2011 , apj , 730 , 15 oppenheimer , b. d. & dav , r. a. 2006 , mnras , 373 , 1265 oppenheimer , b. d. & dav , r. a. 2009 , mnras , 395 , 1875 oppenheimer , b. d. , dav , r. , katz , n. , kollmeier , j. a. , & weinberg , d. h. 2012 , mnras , 420 , 829 oppenheimer , b. d. , & schaye , j. 2013 , arxiv:1302.5710 proux , c. , kulkarni , v. p. , meiring , j. , et al .
2006 , a&a , 450 , 53 prochaska , j. x. , chen , h .- w . , howk , j. c. , weiner , b. j. , & mulchaey , j. 2004 , apj , 617 , 718 prochaska , j. x. , omeara , j. m. , herbert - fort , s. , et al . 2006 , apjl , 648 , l97 rahmati , a. , pawlik , a. h. , raievi , m. , & schaye , j. 2013 , mnras , 431 , 2261 rush , b. , malkan , m. a. , fink , h. h. , & voges , w. 1996 , apj , 471 , 190 savage , b. d. , lehner , n. , wakker , b. p. , semback , k. r. , & tripp , t. m. 2005 , apj , 626 , 776 schaye , j. 2001 , apj , 559 , 507 schaye , j. , aguirre , a. , kim , t .- s . , theuns , t. , rauch , m. , & sargent , w.l.w .
2003 , apj , 596 , 768 schaye , j. , carswell , r. f. , & kim , t .- s .
2007 , mnras , 379 , 1169 schirber , m. , miralda - escud , j. , & mcdonald , p. 2004 , apj , 610 , 105 scott , j. , bechtold , j. , dobrzycki , a. , & kulkarni , v. p. 2000 , apjs , 130 , 67 scott , j. e. , kriss , g. a. , brotherton , m. , et al .
2004 , apj , 615 , 135 shull , j. m. , roberts , d. , giroux , m. l. , penton , s. v. , & fardal , m. a. 1999 , aj , 118 , 1450 shull , j. m. , france , k. , danforth , c. w. , smith , b. , & tumlinson , j. 2010 , apj , 722 , 1312 shull , j. m. , stevans , m. , & danforth , c. w. 2012 , apj , 752 , 162 simcoe , r. a. , sargent , w. l. w. , & rauch , m. 2004 , apj , 606 , 92 simcoe , r.a . ,
sargent , w.l.w . ,
rauch , m. , & becker , g. 2006 , apj , 637 , 648 smith , b. d. , hallman , e. j. , shull , j. m. , & oshea , b. w. 2011 , apj , 731 , 6 steidel , c. c. , hunt , m. p. , shapley , a. e. , et al .
2002 , apj , 576 , 653 steidel , c. c. , erb , d. k. , shapley , a. e. , pettini , m. , reddy , n. , bogosavljevi , m. , rudie , g. c. , & rakic , o. 2010 , apj , 717 , 289 stocke , j. t. , keeney , b. a. , danforth , c. w. , et al .
2013 , apj , 763 , 148 telfer , r. c. , kriss , g. a. , zheng , w. , davidsen , a. f. , & tytler , d. 2002 , apj , 579 , 500 tepper - garcia , t. , richter , p. , schaye , j. , booth , c. m. , dalla vecchia , c. , theuns , t. , & wiersma , r. p. c. 2011 , mnras , 413 , 190 tepper - garca , t. , richter , p. , schaye , j. , booth , c. m. , dalla vecchia , c. , & theuns , t. 2012 , mnras , 425 , 1640 tepper - garcia , t. , richter , p. , & schaye , j. 2013 , arxiv:1303.6281 tescari , e. , viel , m. , dodorico , v. , cristiani , s. , calura , f. , borgani , s. , & tornatore , l. 2011 , mnras , 411 , 826 thom , c. & chen , h .- w .
2008b , apjs , 179 , 37 tilton , e. m. , danforth , c. w. , shull , j. m. , & ross , t. l. 2012 , apj , 759 , 112 tripp , t. m. , semback , k. r. , bowen , d. v. , savage , b. d. , jenkins , e. b. , lehner , n. , & richter , p. 2008
, apjs , 177 , 39 tripp , t. m. , meiring , j. d. , prochaska , j. x. , et al .
2011 , science , 334 , 952 tumlinson , j. , thom , c. , werk , j. k. , et al .
2011 , science , 334 , 948 weedman , d. w. 1977 , ara&a , 15 , 69 wiersma , r. p. c. , schaye , j. , dalla vecchia , c. , booth , c. m. , theuns , t. , & aguirre , a. 2010 , mnras , 409 , 132 wiersma , r. p. c. , schaye , j. , & theuns , t. 2011 , mnras , 415 , 353 worseck , g. , & wisotzki , l. 2006 , a&a , 450 , 495 worseck , g. , fechner , c. , wisotzki , l. , & dallaglio , a. 2007 , a&a , 473 , 805 zuo , l. 1992 , mnras , 258 , 36 | we model the time - dependent evolution of metal - enriched intergalactic and circumgalactic gas exposed to the fluctuating radiation field from an active galactic nucleus ( agn ) .
we consider diffuse gas densities ( @xmath0 ) exposed to the extra - galactic background ( egb ) and initially in thermal equilibrium ( @xmath1 k ) .
once the proximate agn field turns on , additional photo - ionisation rapidly ionises the @xmath2 and metals .
the enhanced agn radiation field turns off after a typical agn lifetime ( @xmath3 myr ) and the field returns to the egb intensity , but the metals remain out of ionisation equilibrium for timescales that can significantly exceed @xmath4 .
we define this phase as the agn proximity zone `` fossil '' phase and show that high ionisation stages ( e.g. @xmath5 , @xmath6 , @xmath7 ) are in general enhanced , while the abundances of low ions ( e.g. @xmath8 , @xmath9 , @xmath10 ) are reduced .
in contrast , @xmath2 re - equilibrates rapidly ( @xmath11 ) owing to its low neutral fraction at diffuse densities .
we demonstrate that metal column densities of intervening gas observed in absorption in quasar sight lines are significantly affected by delayed recombination for a wide range of densities , metallicities , agn strengths , agn lifetimes , and agn duty cycles . as an example
, we show that a fossil zone model can simultaneously reproduce the observed @xmath6 , @xmath10 , @xmath2 , and other metal columns of the @xmath12 pg1206 + 259 absorption system observed by @xcite using a single , @xmath13 k phase model . at low redshift
even moderate - strength agn that are off for 90% of the time could significantly enhance the high - ion metal columns in the circum - galactic media of galaxies observed without active agn .
fossil proximity zones may be particularly important during the quasar era , @xmath14 .
indeed , we demonstrate that at these redshifts a large fraction of the metal - enriched intergalactic medium may consist of out - of - equilibrium fossil zones .
agn proximity zone fossils allow a whole new class of non - equilibrium solutions that may be applicable to a large fraction of observed metal absorbers and which could potentially change the inferred physical conditions and masses of diffuse gases .
[ firstpage ] atomic processes ; plasmas ; galaxies : formation ; intergalactic medium ; quasars : absorption lines ; seyfert ; cosmology : theory ; |
dark matter particles with dominant interaction on electrons have been considered in literature @xcite . in particular , from a phenomenological point of view , dark matter ( dm ) candidates with electron interactions can offer possible sources for the 511 kev positron annihilation line observed from the galactic bulge @xcite .
these candidates can be either light ( mev scale ) @xcite or heavy ( gev or larger scale ) @xcite .
they are expected to interact with electrons both through neutral light ( mev scale ) u or z bosons or through heavy charged mediators @xmath1 ( which can eventually be nearly degenerate with @xmath0 ) @xcite .
recently data collected by some accelerator experiments have been analyzed in terms of a @xmath2 mev neutral boson which couples to quarks with flavour changing transition : @xmath3 @xcite .
other results showing some resonances at energies lower than the two - muon @xcite and the two - pion @xcite disintegration thresholds have been associated with a goldstone neutral boson of @xmath4 mev mass . moreover ,
some excess has been achieved in dedicated experiments on low energy nuclear reactions searching for possible @xmath5 pairs driven by the presence of a neutral boson with a mass around 10 mev @xcite .
let us remark that in the frameworks where the mediator is either a @xmath6 charged boson or a neutral boson providing a flavour changing transition among quarks the elastic scatterings of the dm candidate @xmath7 particles on nuclei would be either forbidden or suppressed ; hence , the scattering on electrons would remain the unique possibility for the direct detection of the @xmath7 particles .
on the other hand , from a pure theoretical point of view , it is also conceivable that the mediator of the dm particle interactions can be coupled only to the lepton sector of the ordinary matter .
thus , in this case the dm particles can just interact with electrons and can not with nuclei .
this is suggested in ref .
@xcite for the u boson and can also be the case of some extensions @xcite or from @xmath8 ^ 4 $ ] quartification @xcite .
] of the standard model providing a quark - lepton discrete symmetry @xmath9 . in these latter models , leptons ( as well as quarks )
are assumed to have three `` leptonic ( @xmath10 ) colours '' and to interact through the gauge group @xmath11 , analogously as the qcd colour group @xmath12 .
moreover , at some high energy scale a symmetry breaking @xmath13 is expected , giving high mass to the `` exotic '' leptonic degree of freedom and leaving light the `` standard '' leptons @xcite . in these scenarios ,
the heavy `` exotic '' leptonic degree of freedom provides both heavy charged @xmath14 fermions , which are expected to be confined into exotic _
leptonic hadrons _ by the unbroken gauge group @xmath15 @xcite , and heavy neutrinos @xcite ; hence , they can be considered as dark matter candidates with dominant interaction on electrons . moreover , it is worth to note that other possibilities can exist .
for example , supersymmetric ( susy ) theories can offer configurations in the general susy parameter space where the lightest supersymmetric particle ( lsp ) has an interaction with electron dominant with respect to that with quark .
these dm candidate particles can be directly detected only through their interaction with electrons in the detectors of a suitable experiment , while they are lost by experiments based on the rejection of the electromagnetic component of the experimental counting rate . in the present paper this kind of dm candidates
are investigated , some theoretical arguments are developed and related phenomenological aspects are discussed . in particular , the impact of these dm candidates will also be discussed in a phenomenological framework on the basis of the 6.3 @xmath16 c.l .
dama / nai model independent evidence for particle dark matter in the galactic halo @xcite .
we remind that various corollary analyses , considering some of the many possible astrophysical , nuclear and particle physics scenarios , have been analysed by dama itself both for some wimp / wimp - like candidates and for light bosons @xcite , while several others are also available in literature , such as e.g. refs .
many other scenarios can be considered as well . at present the new second generation dama / libra set - up is running at the gran sasso laboratory .
the practical possibility to detect electron interacting dm candidates ( hereafter @xmath0 with mass @xmath17 and 4-momentum @xmath18 ) is based on the detectability of the energy released in @xmath0 - electron elastic scattering processes ( see fig . [
fg : diagram_we ] ) .
ht ] @xmath19 elastic scattering and definition of the momentum variables in the laboratory frame . in the text
a contact interaction has been assumed ( also see appendix b ) as suitable approximation of the process.,title="fig : " ] generally , these processes are not taken into account in the dm field since the electron is assumed at rest and , therefore , considering the @xmath7 particle velocity @xmath20 km / s , the released energy is of the order of few ev , well below the detectable energy in any considered detector in the field .
however , the electron is bound in the atom and , even if the atom is at rest , the electron can have not negligible momentum , @xmath21 .
for example , the bound electrons in nai(tl ) offer a probability equal to @xmath22 to have @xmath23 0.5 mev / c ; such a probability is quite small , but not zero .
hence , interactions of @xmath7 particles with these high - momentum electrons in an atom at rest can give rise to detectable signals in suitable detectors .
in particular , after the interaction the final state can have beyond the scattered @xmath7 particle either a prompt electron and an ionized atom or an excited atom plus possible x - rays / auger electrons .
therefore , the process produces x - rays and electrons of relatively low energy , which are mostly contained with efficiency @xmath24 in a detector of a suitable size .
thus , the total detected energy , @xmath25 ( where @xmath26 , @xmath27 , @xmath28 and @xmath29 are the time components of the respective 4-vectors in the laboratory frame , see fig .
[ fg : diagram_we ] ) , can be evaluated considering the energy conservation in the centre of mass ( cm ) frame of the @xmath30 system . defining @xmath31 as the velocity of the cm frame with the respect to the laboratory frame and @xmath32 lorentz boost factor , one can write the energies of the electron before and after the scattering by using the variables in the cm frame through the lorentz transformations : @xmath33 since we are dealing with elastic scattering , @xmath34 and @xmath35 , so that , by subtraction , one obtains : @xmath36 where @xmath37 is the angle between @xmath38 and @xmath39 , @xmath40 is the angle between @xmath38 and @xmath41 and @xmath42 .
therefore , fixing the input momenta of the @xmath7 particle ( @xmath43 ) and of the electron ( @xmath44 ) , the maximum detected energy is given by : @xmath45 .
few examples of the dependence of @xmath46 on the @xmath7 mass are given in fig .
[ fg : emax ] as function of the electron s momentum and of the @xmath7 velocities for head - on collisions ( @xmath47 ) . the fig .
[ fg : emax ] also points out that @xmath7 particles with @xmath17 larger than few gev can provide sufficient energy to be detected in a suitable detector . [ !
ht ] , on the @xmath7 mass for electron s momenta of 0.1 , 1 and 5 mev / c , for @xmath48 ranging in the interval @xmath49 and for head - on collisions ( @xmath47).,title="fig : " ] it is interesting to explore two limit cases ( remind that owing to the typical @xmath7 velocities , @xmath50 and @xmath51 ; hereafter @xmath52 ) : ) @xmath53 kev , that is target nearly at rest nearly at rest : @xmath54 , that is one gets the formula describing for example the wimp - nucleus elastic scattering . ] : @xmath55 ev .
@xmath56 kev ; in this case one obtains @xmath57 , @xmath58 and , therefore , @xmath40 is also the angle between @xmath44 and @xmath43 .
hence : @xmath59 .
this is the case of interest for the direct detection ; in fact , for @xmath17 larger than few gev @xmath60 is larger than the maximum momentum of a bound electron in the atom due to the finite size of the nucleus ( @xmath61 mev in iodine ) . in conclusion ,
@xmath7 particles with mass @xmath62 few gev , interacting on bound electrons with momentum up to @xmath63 few mev / c ( see case b ) , can provide signals in the kev energy region detectable by low background and low energy threshold detectors , such as those of dama / nai ( see later ) .
the differential cross section for @xmath0 - electron elastic scattering can be written as : @xmath64 there @xmath65 is the averaged squared matrix element and @xmath66 is the relative velocity between @xmath0 and the electron . integrating over @xmath67 and over the @xmath68 solid angle and considering that @xmath69 , one can write : @xmath70 the heaviside theta function defines the domain of the differential cross section .
it is useful in the following to define the @xmath7 cross section on the electron at rest ( @xmath71 ) ; thus , one can write : @xmath72 where @xmath73 and @xmath74 . in the following , for simplicity , we define @xmath75 , then @xmath76 .
let us now introduce in the previous evaluations the momentum distribution of the electrons in the atom , @xmath77 ( see appendix a ) . in particular , from eq .
( [ eq : sv33 ] ) that is for a fixed @xmath44 value one can write for the atomic case : @xmath78 introducing the @xmath79 definition and replacing @xmath80 with its expression , it is possible to write for the relevant case of direct detection ( @xmath81 ) : @xmath82 dp ; \label{eq : sv4}\ ] ] here the polar axis has been chosen in the direction of @xmath43 .
the integration over @xmath83 simply gives @xmath84 considering that @xmath65 does not depend on @xmath83 and that atoms with full shells ( as @xmath85 and @xmath86 ) have isotropic distributions @xmath87 .
the expected interaction rate of @xmath7 particle impinging on the electrons of an atom can be derived as : @xmath88 where : i ) @xmath89 with @xmath90 local halo density and @xmath91 fractional amount of @xmath7 density in the halo ; ii ) @xmath92 is the @xmath7 velocity ( @xmath48 ) distribution in the earth frame ; iii ) @xmath93 is the electron s number density in the target material . in the reasonable hypothesis that @xmath79 does not depend on @xmath94 , the integrand in eq .
( [ eq : rate ] ) can be evaluated considering that : @xmath95 where @xmath96 is the minimal @xmath7 particle velocity in order to provide an energy @xmath97 released in the detector .
the matrix element @xmath98 as well as @xmath79 in eq .
( [ eq : rate2 ] ) can generally depend on @xmath21 and @xmath48 .
thus , in order to evaluate it , it is necessary to consider a specific particle interaction model ( see appendix b ) . for simplicity , we will consider a 4-fermion contact interaction ( e.g. a mediator with mass larger than many mev , neglecting the 4-momentum transferred into the propagator ) .
thus , for the cases of pure @xmath99 and pure scalar interactions which are addressed in the following one gets : @xmath100 .
other interaction models are possible and can be investigated in the future .
it is worthwhile to stress that although the calculations are made for the @xmath101 and for the scalar 4-fermion contact interactions
same results can be achieved for any kind of dm candidate interacting with electrons and with cross section @xmath79 having a weak dependence on @xmath21 and @xmath48 , that is @xmath102 .
finally , the expected interaction rate can be written as : @xmath103 where pointing out the time dependence of @xmath92 we have introduced the useful function : @xmath104 here roughly @xmath105 june and @xmath106 with @xmath107 yr . the cut - off of the halo escaping velocity is included into the @xmath92 function distribution .
therefore , the expected counting rate accounting for the energy resolution of the detector can be written as : @xmath108 where @xmath109 and @xmath110 are the unmodulated and the modulated part of the expected signal , respectively .
the @xmath111 kernel generally has a gaussian behaviour .
finally , we note that since @xmath17 is larger than few gev ( so that @xmath112 ) the expected counting rate has a simple dependence upon @xmath113 and @xmath17 ; therefore , the ratio @xmath114 is a normalization factor of the expected energy distribution . the momentum distribution of the electrons in nai(tl ) , @xmath87 , has been depicted in fig .
[ fg : fig5q]a ) ; it has been calculated from the corresponding compton profile , @xmath115 , reported in ref .
for this purpose , due to the isotropic distributions of na@xmath116 and i@xmath117 ( ions with full shells ) the relation @xmath118 has been used @xcite . at high momentum
the @xmath87 function follows the hydrogenic behaviour of the @xmath119 internal shell of the iodine atom : @xmath120 with @xmath121 kev .
h ] ( solid black line ) for nai(tl ) and @xmath122 and @xmath123 for @xmath124 kev in the considered halo model , a5 of ref .
@xcite ; see also text .
the functions @xmath122 and @xmath123 are in arbitrary units .
b ) behaviours of @xmath125 for nai(tl ) at three different values of the released energy : @xmath126 3 , 6 and 12 kev in the considered halo model , a5 of ref .
@xcite ; they show as the main contribution to the counting rate in nai(tl ) detectors with energy threshold at 2 kev comes from electrons with momenta around few mev / c.,title="fig : " ] ( solid black line ) for nai(tl ) and @xmath122 and @xmath123 for @xmath124 kev in the considered halo model , a5 of ref .
@xcite ; see also text .
the functions @xmath122 and @xmath123 are in arbitrary units .
b ) behaviours of @xmath125 for nai(tl ) at three different values of the released energy : @xmath126 3 , 6 and 12 kev in the considered halo model , a5 of ref .
@xcite ; they show as the main contribution to the counting rate in nai(tl ) detectors with energy threshold at 2 kev comes from electrons with momenta around few mev / c.,title="fig : " ] as an example , in fig .
[ fg : fig5q]a ) the behaviours of @xmath127 , @xmath128 and @xmath87 are compared as function of the electron s momentum , @xmath21 , for nai(tl ) as target material and for the given released energy : @xmath129 kev . in this figure as template
the considered halo model is the a5 model of ref .
@xcite , that is a nfw halo model with local velocity equal to 220 km / s and density equal to the maximum value ( @xmath90 = 0.74 gev @xmath130 ) .
h ] interactions with electrons for the scenario given in the text ; the solid line gives the behaviour of the unmodulated part of the expected signal , @xmath109 , while the dashed line is the behaviour of the modulated part , @xmath110 .
in this example the normalization factor is @xmath131 pb / gev .
the vertical line indicates the energy threshold of the dama / nai experiment.,title="fig : " ] it is possible to see that due to the behaviour of the momentum distribution of the electrons , @xmath87 , at high @xmath21 and due to the behaviour of the @xmath132 function at low @xmath21 ( related to the @xmath92 behaviour at high velocity ) the main contribution to the counting rate in nai(tl ) detectors with energy threshold at 2 kev comes from electrons with momenta around few mev / c ( see fig .
[ fg : fig5q]b ) .
it is worthwhile to note that similar behaviours can also be obtained by using other choices of the halo model . finally , an example of the shapes of expected energy distributions in nai(tl ) due to @xmath7 interactions with electrons for the a5 halo model ( a nfw halo model with local velocity equal to 220 km / s and density equal to the maximum value , see ref .
@xcite ) is reported in fig .
[ fg : fig6q ] . in this example
the normalization factor is @xmath131 pb / gev .
the 6.3 @xmath16 c.l . model independent evidence for dark matter particles in the galactic halo achieved over seven annual cycles by dama / nai @xcite ( total exposure @xmath133 kg @xmath134 days ) can also be investigated for the case of an electron interacting dm candidate ( in addition to the other corollary quests already mentioned in the previous footnote 4 ) . in the analysis presented here , the same dark halo models and related parameters given in table vi of ref .
@xcite have been used ; the related dm density is given in table vii of the same reference . moreover , here @xmath93 = 2.6 @xmath134 10@xmath135 kg@xmath136 and the halo escaping velocity has been taken equal to 650 km / s .
the results are calculated by taking into account the time and energy behaviours of the _ single - hit _ experimental data through the standard maximum likelihood method , where @xmath137 is the number of events collected in the @xmath138-th time interval , by the @xmath139-th detector and in the @xmath60-th energy bin .
@xmath137 follows a poissonian distribution with expectation value @xmath140 m_j \delta t_i \delta e \epsilon_{jk}$ ] .
the unmodulated and modulated parts of the signal , @xmath141 and @xmath142 , respectively , are here functions of the only free parameter of the fit : the @xmath143 ratio .
the b@xmath144 is the background contribution ; @xmath145 is the detector running time during the @xmath138-th time interval ; @xmath146 is the overall efficiency and @xmath147 is the detector mass . ] .
in particular , they are presented in terms of the allowed interval of the @xmath143 parameter , obtained as superposition of the configurations corresponding to likelihood function values _
distant _ more than @xmath148 from the null hypothesis ( absence of modulation ) in each one of the several ( but still a very limited number ) of the considered model frameworks .
this allows us to account for at least some of the existing theoretical and experimental uncertainties ( see e.g. in ref .
@xcite and in literature ) . for these scenarios the dama / nai annual modulation data
gives for the considered @xmath7 candidate : @xmath149
pb / gev @xmath150 pb / gev at @xmath148 from null hypothesis . in particular , fig .
[ fg : reg ] shows the dama / nai region allowed in the ( @xmath151 vs @xmath17 ) plane for the same dark halo models and related parameters described in ref .
@xcite .
ht ] vs @xmath17 ) plane for the same dark halo models and related parameters described in ref .
the region encloses configurations corresponding to likelihood function values
_ distant _ more than @xmath148 from the null hypothesis ( absence of modulation ) .
we note that , although the mass region in the plot is up to 2 tev , @xmath7 particles with larger masses are also allowed.,title="fig : " ] we would like to stress that although the above mentioned calculations have been made for the @xmath101 and for the scalar 4-fermion contact interactions the results given here hold for every kind of dm candidate interacting with electrons and with cross section @xmath79 having a weak dependence on @xmath21 and @xmath48 , that is @xmath102 ; in such a case , the dama / nai annual modulation data gives : @xmath152 pb / gev @xmath153 pb / gev at @xmath148 from null hypothesis .
let us now comment some phenomenological implications about the possible mediator of the interaction ( hereafter u boson ) .
the hypothesis of 4-fermion contact interaction still holds for u boson masses , @xmath154 , larger than the transferred momentum ( @xmath155 mev ) . in the pure @xmath101 and pure scalar scenario ,
the cross section is given by ( see appendix b ) : @xmath156 the effective coupling constant , @xmath157 , depends on the couplings , @xmath158 and @xmath159 , of the u boson with the electron and the @xmath7 particle , respectively .
we note that limits on @xmath158 have been achieved by the experimental constraints on the possible u boson coupling to electron arising from the @xmath160 measurements : @xmath161 @xcite .
moreover , more restrictive limits have been obtained under the assumption of universality ( @xmath162 ) by considering the @xmath163 and @xmath164 scattering data : @xmath165 @xcite .
the dama / nai allowed region of fig .
[ fg : reg ] requires values of @xmath158 well in agreement with these experimental upper limits .
in fact , from fig .
[ fg : reg ] and reminding that @xmath91 and @xmath166 few gev ( see above ) , we obtain that @xmath167 pb . requiring that the theory remains perturbative ( that is , @xmath168 ) and for @xmath169 mev , the values of @xmath158 allowed by dama / nai data are ( see eq .
( [ eq : propa ] ) ) : @xmath170 , in agreement with the experimental upper limits . [ ! hb ]
constrain @xcite considering that @xmath91 and that the theory is perturbative ( @xmath168 ) .
there u boson with @xmath154 masses in the sub - gev range required by the analyses of ref .
@xcite is well allowed for a large interval of @xmath17.,title="fig : " ] more in general , considering the limit on @xmath158 from @xmath160 data and the obtained lower bound @xmath171 pb / gev from the dama / nai data , the allowed u boson masses are : @xmath172 , as reported in fig .
[ fg : uboson ] .
there u boson with @xmath154 masses in the sub - gev range required by the analyses of ref .
@xcite is well allowed for a large interval of @xmath17 .
in this paper , the scenario of a dm particle @xmath0 with dominant interaction with electrons has been investigated .
this candidate can be directly detected only through its interaction with electrons in suitable detectors .
theoretical arguments have been developed and related phenomenological aspects have been discussed . in particular
, the impact of these dm candidates has also been analysed in a phenomenological framework on the basis of the dama / nai data . for the considered dark halo models the dama / nai data support for the @xmath7 candidate : @xmath149
pb / gev @xmath150 pb / gev at @xmath148 from null hypothesis . allowed regions for the characteristic phenomenological parameters of the model have been presented .
the obtained allowed interval for the mass of the possible mediator of the interaction is well in agreement with the typical requirements of the phenomenological analyses available in literature .
finally , we further remind that the u boson interpretation is not the unique one since , for example , there are domains in general susy parameter space where lsp - electron interaction can dominate lsp - quark one .
the inclusive scattering of @xmath7 particle on an atom @xmath173 is here analyzed : where @xmath174 denotes the final state of the atom .
the cross section of the process is obtained by summing over the possible contributions of all the @xmath174 final states : @xmath175 here @xmath176 is the transition amplitude when the final state is @xmath174 . since it has been assumed that the interaction of @xmath7 with the electrons is dominant , we can use a full set of electronic plane wavefunctions , @xmath177 , and rewrite : @xmath178 @xmath179 therefore : @xmath180 where @xmath181 is the transition amplitude for free electron @xmath182 elastic scattering and @xmath183 is the matrix element reported in eq .
( [ eq : d1 ] ) . since @xmath174 is whatever final state : @xmath184 ; therefore , eq . ( [ eq : a1 ] ) can be written as : @xmath185 where @xmath186 is the momentum distribution function of the electrons in the atom @xmath173 .
finally , we can deduce @xmath187 , where @xmath188 is the @xmath30 elastic scattering cross section given in eq .
( [ eq : d1 ] ) .
in the following we consider the elastic scattering of the @xmath7 fermion on electron by using a fermi - like 4-fermion contact interaction .
where : @xmath190 \left [ \bar{u}_{\chi^0}(k ) \gamma^{\nu } ( g_v+g_a \gamma^5 ) u_{\chi^0}(k ' ) \right ] \label{eq : va2}\ ] ] @xmath191 \left [ \bar{u}_{e}(p ) \gamma_{\nu } ( c_v+c_a \gamma^5 ) u_{e}(p ' ) \right ] \ ; .
\label{eq : va3}\ ] ] let us focus just on eq .
( [ eq : va2 ] ) , since eq .
( [ eq : va3 ] ) has the same structure .
one can write : @xmath192 \nonumber \\ & = & t^{aa}+t^{va}+t^{av}+t^{vv } \label{eq : va4}\end{aligned}\ ] ]
the four terms can be explicited as : @xmath193 \label{eq : va5}\ ] ] @xmath194 \label{eq : va6}\ ] ] @xmath195 \label{eq : va7}\ ] ] @xmath196 \label{eq : va8}\ ] ] by using trace theorems one gets : @xmath197=2 g_a^2(k'^{\mu}k^{\nu}+k'^{\nu}k^{\mu}-k'kg^{\mu\nu}- m_{\chi^0}^2 g^{\mu\nu } ) \label{eq : va9}\ ] ] @xmath198 = 2 g_v^2 ( k'^{\mu}k^{\nu}+k'^{\nu}k^{\mu}-k'kg^{\mu\nu } + m_{\chi^0}^2 g^{\mu\nu } ) \label{eq : va10}\ ] ] @xmath199 \label{eq : va11}\ ] ] @xmath200=t^{av } \label{eq : va12}\ ] ] @xmath201 in the case of @xmath99 interaction ( @xmath205 and @xmath206 ) the matrix element is : @xmath207 \label{eq : v - a}\end{aligned}\ ] ] knowing that @xmath7 is not relativistic ( see text ) , one obtains : @xmath208 and @xmath209 ; moreover for @xmath210 one has @xmath211 , giving : @xmath212 where the fermi effective coupling constant is : @xmath213 . for this particular case ,
the dependence on @xmath48 can be neglected , while the dependence on @xmath21 are included in : @xmath214 .
similarly as above , one has : @xmath215 @xmath216 \left [ \bar{u}_{\chi^0}(k ) ( g_s+ig_p \gamma^5 ) u_{\chi^0}(k ' ) \right ] \label{eq : sp2}\ ] ] @xmath217 \nonumber \\ & = & t^{ss}+t^{sp}+t^{ps}+t^{pp } \label{eq : sp3}\end{aligned}\ ] ] there : @xmath218=2g_s^2(k'k+m_{\chi^0}^2 ) \label{eq : sp4}\ ] ] @xmath219=2g_p^2(k'k - m_{\chi^0}^2 ) \label{eq : sp5}\ ] ] @xmath220 \label{eq : sp6}\ ] ] @xmath221 = 0 \label{eq : sp7}\ ] ] hence : @xmath222= 2(g_+k'k+g_-m_{\chi^0}^2 ) , \label{eq : sp8}\ ] ] where @xmath223 and @xmath224 .
finally : @xmath225 \label{eq : sp9}\ ] ] in the particular pure scalar case ( @xmath226 ) one obtains : @xmath227\left[(p'p)+m_e^2 \right]\simeq 8 g^2_s m_{\chi^0}^2 \left[p'_0p_0-\vec{p'}\vec{p}+m_e^2 \right ] \label{eq : sp92}\ ] ] thus , considering the momentum distribution of atomic electron , for @xmath210 practically @xmath228 and , therefore : @xmath229 y. ascasibar , p. jean , c. boehm and j. knoedlseder , mon . not .
368 ( 2006 ) 1695 ; c. jacoby and s. nussinov , jhep 05 ( 2007 ) 017 .
finkbeiner and n. weiner , phys .
d 76 ( 2007 ) 083519 . m. pospelov and a. ritz , phys . lett .
b 651 ( 2007 ) 208 , hep - ph/0703128 .
p. fayet , phys .
d 75 ( 2007 ) 115017 .
j. knodlseder et al .
astrophys . 513
( 2005 ) 441 ; p. jean et al .
l55 ( 2003 ) 407 ; j. knodlseder et al . ,
l457 ( 2003 ) 411 . c. boehm and y. ascasibar , phys
d 70 ( 2004 ) 115013 ; g. weidenspointner et al .
, astro - ph/0702621 . b. tatischeff and e. tomasi - gustafsson , arxiv:0710.1796 and arxiv:0710.1798 .
deshpande , g.eilam and j. jiang , phys .
b 632 ( 2006 ) 212 ; d.s .
gorbunov and v.a .
rubakov , phys .
d 73 ( 2006 ) 035002 ; c.h .
chen et al . , arxiv:0708.0937 .
t. walcher , hep - ph/0111279 .
de boer et al . , j. phys .
g 27 ( 2001 ) l29 ; j. phys .
g 23 ( 1997 ) l85 ; nucl .
b 72 ( 1999 ) 189 ; m. el - nadi and o.e .
badawy , phys .
61 ( 1988 ) 1271 ; k. asakimori et al .
, j. phys .
g 25 ( 1999 ) l133 .
r. foot , h. lew and r.r .
volkas , phys .
d 44 ( 1991 ) 859 .
k.s . babu , e. ma and s. willenbrock , phys .
d 69 ( 2004 ) 051301(r ) ; s.l .
chen and e. ma , mod .
lett . a 19 ( 2004 ) 1267 ; a. demaria , c.i .
low , r.r .
volkas , phys .
d 72 ( 2005 ) 075007 .
r. foot and h. lew , phys .
d 41 ( 1990 ) 3502 ; r. foot , h. lew and r.r .
volkas , phys .
d 44 ( 1991 ) 1531 ; r. foot and r.r .
volkas , phys .
b 645 ( 2007 ) 345 ; k.s . babu , t.w .
kephart , h. pas , arxiv:0709.0765 [ hep - ph ] .
r. bernabei el al . , la rivista del nuovo cimento 26 n.1 ( 2003 ) 1 - 73 .
r. bernabei el al . , int . j. mod
d 13 ( 2004 ) 2127 .
r. bernabei et al .
j. c. 47 ( 2006 ) 263 .
r. bernabei el al . , int . j. mod
. phys . a 22 ( 2007 ) 3155 - 3168 .
r. bernabei et al .
phys . a 21 ( 2006 ) 1445 .
r. bernabei et al . , to appear on eur .
phys . j. c , arxiv:0710.0288 .
a. bottino et al . , phys .
d 67 ( 2003 ) 063519 ; a. bottino et al . , phys . rev .
d 68 ( 2003 ) 043506 .
a. bottino et al .
d 69 ( 2004 ) 037302 .
a. bottino et al . , phys .
b 402 ( 1997 ) 113 ; phys . lett .
b 423 ( 1998 ) 109 ; phys . rev .
d 59 ( 1999 ) 095004 ; phys . rev .
d 59 ( 1999 ) 095003 ; astrop . phys . 10
( 1999 ) 203 ; astrop .
13 ( 2000 ) 215 ; phys .
d 62 ( 2000 ) 056006 ; phys . rev .
d 63 ( 2001 ) 125003 ; nucl . phys .
b 608 ( 2001 ) 461 .
k. belotsky , d. fargion , m. khlopov and r.v .
konoplich , hep - ph/0411093 .
d. smith and n. weiner , phys .
d 64 ( 2001 ) 043502 ; d. tucker - smith and n. weiner , phys . rev .
d 72 ( 2005 ) 063509 .
r. foot , hep - ph/0308254 .
s. mitra , phys .
d 71 ( 2005 ) 121302(r ) .
drobyshevski et al . , arxiv:0704.0982 e.m .
drobyshevski , arxiv:0706.3095 c. arina and n. fornengo , arxiv:0709.4477 a. bottino et al .
, arxiv:0710.0553 p. belli et al .
d 66 ( 2002 ) 043503 .
f. biggs et al . , atomic data and nuclear data tables 16 ( 1975 ) 201 .
d. brusa et al . , nucl .
& meth . a 379 ( 1996 ) 167 .
r. ribberfors et al .
a 26 ( 1982 ) 3325 ; phys .
b 12 ( 1975 ) 2067 . | some extensions of the standard model provide dark matter candidate particles which can have a dominant coupling with the lepton sector of the ordinary matter .
thus , such dark matter candidate particles ( @xmath0 ) can be directly detected only through their interaction with electrons in the detectors of a suitable experiment , while they are lost by experiments based on the rejection of the electromagnetic component of the experimental counting rate .
these candidates can also offer a possible source of the 511 kev photons observed from the galactic bulge . in this paper
this scenario is investigated .
some theoretical arguments are developed and related phenomenological aspects are discussed . allowed intervals and regions for the characteristic phenomenological parameters of the considered model and of the possible mediator of the interaction
are also derived considering the dama / nai data .
# 1#2#3#4#1 # 2 ( # 3 ) # 4 * rom2f/2007/19 + * * to appear on phys .
rev .
d + * = 0.65 cm * investigating electron interacting dark matter + * r.bernabei , p.belli , f.montecchia , f.nozzoli _ dip .
di fisica , universit di roma tor vergata " and infn , sez .
roma tor vergata " , i-00133 rome , italy _
f.cappella , a.incicchitti , d.prosperi _ dip .
di fisica , universit di roma la sapienza " and infn , sez .
roma , i-00185 rome , italy _
r.cerulli _ laboratori nazionali del gran sasso , infn , assergi , italy _
c.j.dai , h.l.he , h.h.kuang , j.m.ma , x.h.ma , x.d.sheng , z.p.ye , r.g.wang , y.j.zhang _ ihep , chinese academy , p.o .
box 918/3 , beijing 100039 , china _ _ keywords : _ dark matter ; underground physics _ pacs numbers : _ 95.35.+d |
what is the maximum temperature where the bulk of qcd thermodynamics can be described in hadronic terms , with no explicit reference to the underlying quarks and gluons ? in this contribution
we analyze to what extent can one describe the qcd thermodynamics from the hadronic spectrum .
we thus hope to identify the nature of non - hadronic precursors of the crossover to the quark - gluon plasma in terms of low temperature partonic expansions @xcite ( see @xcite for a review ) .
the relation between spectrum and partition function provides the thermodynamic free energy @xmath0 @xmath1 the example of a hydrogen gas in qed illustrates the situation and motivates the discussion regarding the completeness of states which ultimately are interacting protons and electrons . at low temperatures , states are molecular , i.e. @xmath2 states which behave as compact and structureless .
when we heat up the system a rovibrational spectrum of states @xmath3 emerges and eventually the molecule is disociated into atoms @xmath4 . at higher temperatures ,
atoms are excited @xmath5 and eventually ionized into a plasma of @xmath6 and @xmath7 in the continuum where the @xmath6 and @xmath7 constituents becomes manifest .
clearly , neutral states ( atomic or molecular ) are not effectively complete at all temperatures and residual interactions corresponding to background non - resonant scattering become relevant @xcite .
in qcd on the lattice @xmath0 is determined by integrating the trace anomaly , @xmath8 with respect to the temperature and _ assuming _ a reference value , ideally @xmath9 with @xmath10 .
once this condition is imposed we expect that at sufficiently low temperatures all the observables are described in terms of hadronic degrees of freedom ; the quark - gluon underlying constituents are disclosed at very high temperatures .
this quark - hadron duality _ at low temperature _ has been confirmed by the most recent and accurate lattice calculations at finite temperature @xcite are confronted with the hadron resonance gas ( hrg ) , a rather simple system of non - interacting point - like and structureless hadrons @xmath11 here , the sum is over _ all _ hadronic states including spin - isospin and anti - particle degeneracies , @xmath12 for mesons and baryons respectively , @xmath13 is the energy and the cumulative number of states ( degeneracy included ) @xmath14 where @xmath15 are the hadron masses .
the spectrum can be taken either directly from the pdg @xcite or the relativized quark model ( rqm ) refs .
indeed , requiring @xmath16 ^ 2 \sim n_{\rm dat } \pm \sqrt{2 n_{\rm dat } } \ , , \label{eq : chi2}\end{aligned}\ ] ] we get @xmath17 for @xmath18 lattice points @xcite both for pdg and rqm , showing that they can be used to saturate the hadronic spectrum for light quarks provided @xmath19 .
this rqm saturation will be assumed also to take place when heavy quarks are involved in which case the pdg lacks many of the needed states .
this of course suggests to determine _ directly _ @xmath20 from the @xmath21 data , as illustrated in fig .
[ fig : deut ] ( left panel ) .
the partition function and its comparison with hadronic states through the trace anomaly @xmath21 , displays explicitly the connection between spectrum and thermodynamics yielding the statistically significant identification of @xmath22 with @xmath23 using pdg or rqm where only @xmath24 and @xmath25 states contribute .
but , which hadronic states count into the calculation of the thermodynamic properties ? . as argued long ago @xcite , counting hadronic states implicitly averages over some scale , and so states such as the deuteron ( made of @xmath26 ) generate fluctuations in a smaller nuclear scale .
the cumulative number in a given channel @xmath27 with bound states @xmath28 below the threshold @xmath29 is in general ( see e.g. @xcite ) @xmath30 \ , , \end{aligned}\ ] ] which becomes @xmath31/\pi=0 $ ] due to levinson s theorem .
this fluctuation is shown in fig . [
fig : deut ] for the nn channel where @xmath32 , and the deuteron mass is @xmath33 ( @xmath34 ) .
thus , the deuteron does nt count .
this is a general feature of weakly bound states , which for the copious new x , y , z states , might affect the thermodynamics if the hrg was blindly identified with the pdg ( with _ all _ x , y , z s ) .
the linearly growing static energy of two @xmath35 and @xmath36 sources in the fundamental representation of the su(@xmath37)-group placed at a distance @xmath38 is often identified with confinement with a string tension @xmath39 . including the short distance ( perturbative ) coulomb - like behaviour yields the venerable cornell potential for the ground state @xmath40
actually , the mass of a @xmath41 state , becomes unstable at a critical distance @xmath42 when a light @xmath43 pair is created from the vacuum and the heavy - light meson - antimeson @xmath44 channel opens @xmath45 where the weak van der waals like meson exchange interaction in the @xmath46 sector is neglected proceeds as follows . due to spontaneous chiral symmetry breaking
the quarks acquire an additive _ dynamical _ mass @xmath47 contribution to the total mass @xmath48 where @xmath49 is the _ current _ mass . for light @xmath43-mesons @xmath50 whereas for light - heavy ( @xmath51 ) mesons @xmath52 and thus @xmath53 which yields @xmath54 in fair agreement with the lattice estimate @xcite . ] . we have a diabatic crossing structure which turns into an adiabatic avoided crossing when the transition strength @xmath55 is included . in general , one has excited meson states , @xmath56 where @xmath57 and in fig .
[ fig : isgur3 ] we show the resulting adiabatic spectrum based on the rqm @xcite when a mixing strength of @xmath58 is implemented .
much of our understanding of the qcd dynamics at finite temperature is linked to the helmholtz free energy defined as the maximum work the system can exchange at fixed temperature between two heavy @xmath59 at separation @xmath38 . in the confined phase , the correlator between polyakov loops becomes @xcite @xmath60 where , before mixing , one has the crossing among the energy levels up to @xmath61 pair creation . neglecting the avoided crossing , eq .
( [ eq : fave ] ) yields @xmath62 where @xmath63 .
we show in fig .
[ fig : flisgurmesons ] the heavy @xmath64 free energy obtained with eq .
( [ eq : fave2 ] ) by using the spectrum of heavy - light mesons with a charm ( bottom ) quark , and no strangeness , obtained with the isgur model of ref .
@xcite up to @xmath65gev .
we have considered @xmath66 and @xmath67 ( no fit ) . in fig .
[ fig : flisgurmesons ] we also show the polyakov loop computed as @xmath68 the agreement with lattice data is quite good for @xmath69 .
thus the spectrum both for the pdg and the rqm saturates the sum rules at these temperatures by a factor @xmath70 with @xmath71mev , we get good agreement with the lowest temperature lattice point .
this kind of ambiguity comes from renormalization effects . like
in @xcite we assume @xmath72 for @xmath73 .
the recent evaluation of ref .
@xcite , unlike @xcite is based on a different renormalization condition and @xmath74 for the free energy and the role of avoiding crossing will be analyzed elsewhere . ] whereas the finite temperature does not resolve avoided crossings making them effectively diabatic passages .
while much of the effort has been directed towards a description of the hadron - to - quark - gluon crossover , at present there is no understanding on the mechanism , perhaps because confinement itself is not really well understood .
our analysis of the free energy suggests that lowest temperatures where purely hadronic states account for qcd observables is rather large @xmath75 and close to the critical temperature .
e. megias , e. ruiz arriola and l. salcedo , phys.rev.lett . 109 ( 2012 )
e. megias , e. ruiz arriola and l. salcedo , phys.rev .
d89 ( 2014 ) 076006 .
e. ruiz arriola , l. salcedo and e. megias , acta phys.polon .
b45 ( 2014 ) 2407 .
r. rajaraman , mesons in nuclei 3 ( 1979 ) . | we show how the quark free energy can be determined from a string and the hadron resonance gas model with one heavy quark below the de - confinement phase transition .
we discuss the interesting problem of identification of degrees of freedom at increasing temperatures , as well as the relevance of string breaking and avoided crossings . |
in fig . ( [ asw ] ) , we demonstrate that the compressional solitary wave ( sw ) excitation discovered by nesterenko in a chain of non - cohesive beads is also seen in a lattice of springs with two sided interactions .
however , unlike the case of a one sided potential , each compressive sw generated in response to an impulse is accompanied by a corresponding expansion solitary wave ( formed by local stretching of springs ) of the same magnitude but moving in the opposite direction . this anti solitary wave ( asw )
is not sustained by beads interacting with purely repulsive potentials the beads would merely loose contact . in fig .
( [ asw ] ) left panel , we show the sw / asw excitations for ( a)beads , ( b - c ) particles connected by springs . in fig .
( [ asw ] ) , right panel , we plot , @xmath2 , the _ total _ energy carried by the soliton ( after summing over all particles envolved ) versus @xmath0 , the _ total _ momentum , for the leading sw / asw .
note that @xmath1 , where the constant @xmath3 can be viewed as the effective mass of the solitary or antisolitary waves .
inspection of fig .
( [ asw ] ) demonstrates that sw excitations in a lattice of repulsive beads ( black circles ) have the same effective mass @xmath3 as a sw and asw in two - sided springs ( red squares ) . as shown in fig .
( [ asw ] ) left panel ( b - c ) , the leading sw - asw generated in response to an impulse imparted to one of the particles towards the right ( direction of arrow ) is followed by a train of alternating sw - asw s excitations , of progressively smaller magnitudes .
the smaller sw / asw s are generated as the particle that initially imparted the impulse , recoils with its left - over energy .
this process is repeated several times , leading to the generation of the train of smaller excitations . since the speed of propagation depends upon the amplitude , the sw and asw that start propagating together initially ( appearing bounded ) , eventually separate and become clearly distinguishable .
the classical energy - momentum relation @xmath1 satisfied by the sw motivates the interpretation of the solitary wave as a quasi - particle @xcite . for small perturbations ,
the sw can still be treated as a quasi - particle provided the effects of the perturbations accrue gradually such that the sw retains its functional form .
we now apply this adiabatic approximation to derive an effective langevin equation for the sw quasi - particle when the lattice of springs is coupled to a heat bath .
recall first , the langevin equation for a particle of mass @xmath4 undergoing brownian motion in one dimension is @xmath5 here , @xmath6 are the total and kinetic energies respectively , @xmath7 are the dissipation and diffusion coefficients related via the fluctuation dissipation theorem @xmath8 , where @xmath9 is the boltzmann constant .
@xmath10 is a normal random variable with mean 0 and variance 1 , and encapsulates the effects of random fluctuations during the time interval @xmath11 . for a free particle of unit mass moving with speed @xmath12 , @xmath13 and upon substituting in eq .
( [ langevin_particle ] ) , we recover the langevin s equation conventionally expressed as the rate of change of momentum of the particle @xcite .
we now derive an equation analogous to eq .
( [ langevin_particle ] ) for the compressive solitary wave quasi - particle .
let the displacement of a particle from its initial equilibrium position in the continuum limit be @xmath14 .
if we identify the lattice spacing @xmath15 as a characteristic length scale and @xmath16 as an inverse time scale , then the equation of motion for the compressive displacement field @xmath17 in dimensionless units reads @xmath18_{x}= 0 \label{boussinesq}\end{aligned}\ ] ] where subscripts denote partial derivatives with respect to space @xmath19 and time @xmath20 .
( [ boussinesq ] ) is a simplified form of the nesterenko equation @xcite , see appendix a for details .
the first two terms express the rate of change of momentum while the third term represents the force . although the solitary wave solution to eq .
( [ boussinesq ] ) is not exact ( lacking compact support ) , eq .
( [ boussinesq ] ) provides a good approximation while being analytically more tractable especially since we are interested in keeping the non - linear exponent @xmath21 general @xcite .
note that the equation for the asw ( stretching ) is obtained by modifying the third term @xmath22_{x } \rightarrow - [ { ( \phi _ x)}^{\alpha-1}]_{x}$ ] in eq .
( [ boussinesq ] ) . in analogy with the langevin equation for a particle , we model the coupling to a heat bath as the sum of two contributions- an external drag and a random fluctuating force , phenomenologically introduced into the equation of motion as : @xmath18_{x}= -\gamma\left(\phi _ t - \frac{1}{12}\phi _ { txx}\right ) + \nonumber \\
\sqrt{\frac{2\gamma}{\alpha\gamma dt } } \left(\eta(x , t;t+dt ) - \frac{1}{\sqrt{12}}\eta _ x(x , t;t+dt)\right ) \label{boussinesq_1}\end{aligned}\ ] ] where @xmath23 is the dimensionless drag coefficient that couples to the momentum @xmath24 .
it is useful to define a coupling constant @xmath25 as the ratio of potential to thermal energy in terms of which , the dimensionless diffusion coefficient is @xmath26 . the last ( noise ) term on the right of eq .
( [ boussinesq_1 ] ) in conjunction with @xmath27 , satisfies the fluctuation dissipation theorem @xcite . here
, @xmath28 is a gaussian random noise during the time interval @xmath11 with the moments @xmath29 and @xmath30 respectively , where angular brackets denotes ensemble averaging . to study the propagation of the sw in a background of thermal fluctuations , we now make a working assumption based on the quasi - particle approximation to the sw : whenever the energy of the sw , @xmath31 , the sw functional form is unaltered and only its amplitude @xmath32 becomes time - dependent .
the amplitude @xmath32 is the collective variable for the sw quasi - particle and other properties of the solitary wave , such as its energy and momentum may be determined from it .
note , the width of the sw is independent of its amplitude and therefore we do not consider its time dependence @xcite . from eq .
( [ boussinesq ] ) , the conserved energy is @xmath33 and the energy of the sw may be obtained by integrating eq .
( [ hamiltonian ] ) over the width of the sw of order @xmath34 .
( this avoids including the energy of small sw that separate from the main wave ) . using eq .
( [ boussinesq ] ) , the rate of change of energy is , @xmath35 where , @xmath36 is the kinetic part of the energy , @xmath37 the last two terms on the right of eq .
( [ sw_langevin ] ) describe the possible mechanisms of decay of the sw by `` friction '' from the heat bath ( @xmath38 ) and the `` phonon drag '' induced by the thermal motion of the chain @xmath39 .
the first term is the fluctuating part of the energy . in the following ,
we make the assumption ( verified numerically ) that the coupling to the heat bath is more important and therefore , ignore @xmath39 ( see si for details ) . solving for the sw solution from eq .
( [ boussinesq ] ) , we find @xmath40 , where @xmath41 is the functional form of the sw with amplitude @xmath42 , speed @xmath43 and width @xmath44 in units of the lattice spacing , see appendix a , subsection 1 , for details .
the sw energy , kinetic energy , and momentum may now be expressed in terms of the collective variable @xmath42 : @xmath45 and from the the virial theorem , @xmath46 additionally , the solitary wave momentum is @xmath47 here , @xmath48 are constants obtained by integrating over all space @xcite . substituting for @xmath2 and @xmath36 in terms of @xmath42 , we cast eq .
( [ sw_langevin ] ) into the form of an ordinary langevin equation ( with additive noise ) for the collective variable @xmath32 , @xmath49 where , @xmath50 .
( [ soliton_langevin ] ) is the central result of our work whose analytical predictions we derive and test numerically in the next sections .
the first term can be written as @xmath51 , where @xmath52 is a white noise signal ; that is , its correlations are given by @xmath53 . using the fact that the correlations of @xmath54 are described by delta - functions ,
the correlations of @xmath55 can be related to the kinetic energy eq .
( [ ke ] ) , which can be replaced by @xmath56 .
taking the expectation value ( ensemble average ) of eq .
( [ soliton_langevin ] ) , we find @xmath57 where , owing to the noise term @xmath58 ( which acts between times @xmath59 ) and @xmath60 ( which is a solution at time @xmath20 ) being statistically independent , the expectation value @xmath61 .
consequently , the solitary wave amplitude decays as @xmath62 where , @xmath63 is the initial solitary wave amplitude .
note , the effective damping rate @xmath64 is independent of inverse temperature @xmath65 but rescales with the exponent of the non - linear potential @xmath21 . similarly , we solve for the variance of the solitary wave amplitude or equivalently , the variance in the square root of energy .
re - defining , @xmath66 , we solve for the variance in the solitary wave amplitude by first evaluating the differential @xmath67=a^2(t+dt)-a^2(t)$ ] , by substituting @xmath68 from eq .
( 7 ) and retaining terms to order @xmath69@xcite , @xmath70 = -\frac{2\alpha\gamma}{\alpha+2}a^2dt + 2a\sqrt{ddt}\int dx\ \eta(x , t;t+dt)\left(\psi+\frac{1}{\sqrt{12}}\psi _ ( x)\right)+ \nonumber \\
ddt\int\int dx dx ' \left(\psi(x , t)+\frac{1}{\sqrt{12}}\psi _ { x}(x , t)\right)\left(\psi(x',t)+\frac{1}{\sqrt{12}}\psi _ { x'}(x',t)\right).\end{aligned}\ ] ] taking the expectation value , the second term on the right vanishes ( as discussed for the mean ) and using the property that the noise term is delta - correlated in space , we obtain @xmath71 = -\frac{2\alpha\gamma}{\alpha+2}\langle a^2\rangle dt + ddt\int dx \ \left(\psi+\frac{1}{\sqrt{12}}\psi _ { x}\right)^2.\end{aligned}\ ] ] the last term when expanded gives twice the solitary wave kinetic energy @xmath72 , see eq .
( [ ke ] ) , plus an integral @xmath73 , that vanishes by symmetry for the sw solution .
moreover , the sw kinetic energy is related to its total energy via the virial relation @xmath74 .
hence , we obtain the ordinary differential equation correct to order @xmath75 - @xmath76 solving , the differential equation subject to the initial condition @xmath77 and substituting for @xmath78 , we obtain , @xmath79 using eq .
( [ sw_damping ] ) , this may be expressed as @xmath80 using the relation in eq .
( [ sw_e ] ) , we rewrite the above equation as @xmath81 the coefficient @xmath82 reduces to @xmath83 when the energy is not measured in units of @xmath84 , so this expression is analogous to the velocity variance of a brownian particle . note , for large @xmath21 , the sw is effectively one particle wide and thus eq . ( [ var ] ) captures the correct thermal equilibration of the particle energy with the heat bath .
however , for the dynamics of the sw , eq .
( [ var ] ) is only useful as long as the sw is identifiable against the background thermal energy , that is @xmath85 .
( red ) and @xmath86 ( black ) .
_ right _ : the attenuation of the solitary wave as a function of time for various values of @xmath38,@xmath65 , and @xmath21 . when not indicated , @xmath87 and @xmath88 .
the data for these values is shown with green circles ; the other data corresponds to changing one parameter , @xmath86 ( black circles ) , @xmath89 , ( brown squares ) and @xmath90 ( red circles ) compared with the analytic expression in eq .
( [ sw_damping ] ) ( log scale ) represented by solid lines . the initial @xmath91 .
, scaledwidth=50.0% ]
( inverse temperature ) for @xmath92 .
the circles are from numerical simulation while the dashed blue line is a linear fit , giving a slope of 0.11 for @xmath92 , close to the expected value @xmath93 .
the inset shows the kinetic and potential energy approaching thermal equilibrium , where their ratio satisfies the virial relation @xmath94.,scaledwidth=50.0% ] we consider a one dimensional chain consisting of @xmath95 particles each having a mass @xmath4 placed regularly on a lattice with spacing @xmath15 ( spring rest length ) interacting pair - wise with a nearest neighbour interaction @xmath96 , where @xmath97 is the compression / stretching induced during the dynamics .
we model the coupling to the heat bath by numerically integrating eq .
( [ langevin_particle ] ) for each particle using the velocity - verlet algorithm @xcite . in thermal equilibrium
the mean kinetic energy is ke@xmath98 and potential energy is pe@xmath99 , where their ratio satisfies the virial relation , see fig .
( [ dispersion ] ) , inset for @xmath100 . in the following ,
all numerical data is presented in dimensionless units , ensemble averaged over 1000 samples . to extract the equilibrium properties in the thermalized state , we define the longitudinal current density of particles as @xmath101 , and its fourier transform @xmath102 , where @xmath103 is the longitudinal collective mode along the @xmath19-direction .
thus , the corresponding longitudinal current density auto - correlation functions is @xmath104 where the angular brackets denote ensemble averaging over the initial time .
the longitudinal power spectral density is then obtained as the fourier transform of the respective current density auto - correlation functions as , @xmath105 .
the fourier transforms defined above are evaluated using fast fourier transform from simulation data . the sound speeds in fig . ( [ dispersion ] )
correspond to the linear part of the dispersion curves , obtained by projecting the power spectral densities on the frequency ( @xmath106)- wavenumber ( @xmath103 ) plane . in fig .
[ dispersion ] , we plot the sound speed from the slope of the dispersion curves for @xmath100 for a range of @xmath65 . at thermal equilibrium ,
the mean kinetic energy and hence the temperature @xmath107 satisfy the virial relation @xmath108 , where @xmath109 is the average displacement of the particles induced by thermal fluctuations . defining the sound speed @xmath110 as the second derivative of the induced potential energy leads to the relation , @xmath111 @xcite . for @xmath92 , we find @xmath112 , closely matching the linear fit in si fig .
thus , coupling the lattice of non - linear springs that is initially in its state of sonic vacuum ( implying the absence of linear sound ) to a heat bath , leads to hydrodynamical sound modes with a linear sound speed that scales with the temperature of the heat bath @xcite .
note , setting @xmath113 ( harmonic springs ) yields a sound speed that is independent of temperature while the limit @xmath114 yields @xmath115 , a result in agreement with the entropic elasticity for hard sphere colloidal crystals @xcite .
( purple circles ) , @xmath116 ( blue squares ) , @xmath117 ( black circles ) , @xmath118 ( gray squares ) decaying exponentially compared against the analytical expression ( solid curves ) .
the mean decay rate is independent of the temperature @xmath119 .
_ right _ : the numerically obtained varaince of the solitary wave energy for @xmath120 ( blue squares ) , @xmath121 ( black circles ) , @xmath122 ( grey squares ) , @xmath123 ( purple circles ) and @xmath124 ( red squares ) compared against the analytical expression eq .
[ var ] ( solid curves).,scaledwidth=50.0% ] once the lattice reaches thermal equilibrium , we excite a solitary wave ( sw ) by imparting one of the particles an initial energy of order @xmath91 in dimensionless units . in fig .
( [ damping ] ) left panel , we show a snapshot of two sws at the same time , propagating in a background of thermal fluctuations for @xmath125 ( red ) and @xmath86 ( black ) . we see that the sw with lower @xmath21 is wider and moves faster for the given amplitude , in qualitative agreement with the analytic widths @xmath126 and speeds @xmath127 . in fig .
( [ variance ] ) left panel , we plot the numerical data ( symbols ) for the attenuation of the sw amplitude as a function of time for various values of @xmath38 and @xmath21 and we find a very good match to the analytic expression in eq .
( [ sw_damping ] ) ( solid curves ) . for the range of @xmath65 explored ,
we find the damping rate is independent of temperature ( @xmath65 ) but depends on the environmental drag @xmath38 and @xmath21 . in fig .
( [ variance ] ) right panel , we show the increase in the variance of sw amplitude ( or the square root of its energy ) as a function of time for multiple values of @xmath21 , @xmath38 and @xmath65 obtained numerically ( symbols ) and compare them with the complete analytical solution eq.([var ] ) finding good agreement .
notice , the final value of the variance correctly approaches the thermal energy , as expected for a brownian particle .
however , since the solitary wave is a dynamical object that decays under the influence of the external drag , once the solitary wave energy becomes comparable to the background thermal energy , it is no longer meaningful to consider it as a brownian particle .
in this section , we will review the rosenau approximation to the nesterenko solitary wave solution , that is valid for any general non - linear potential @xcite . here
, we adopt as our starting point the lagrangian for a one dimensional chain of identical spheres that are just touching each other , i.e. , in the limit @xmath128 - @xmath129 where , @xmath130 is the displacement of the @xmath131th sphere from its equilibrium position , @xmath132 is the equilibirum lattice spacing and @xmath36 is the spring constant . in order to avoid doing a binomial expansion in powers of @xmath21 , we will define the continuum field variable as @xmath133 where , primes denote derivative with respect to @xmath19 .
we now take the continuum limit , i.e. , @xmath134 and taylor expand the right hand side about @xmath135 : @xmath136 integrating both sides once with respect to @xmath19 , we obtain @xmath137 inverting the differential operator , we obtain @xmath138 thus , in the contiuum limit , the lagrangian becomes @xmath139 by using the euler - lagrange equation , we obtain the equation of motion as @xmath140'=0.\end{aligned}\ ] ] note , @xmath141 here corresponds to the continuum displacement field . the corresponding equation in the strain field @xmath142 reads @xmath143''=0.\label{ross}\end{aligned}\ ] ] upon substituting @xmath144 for the compressive sw or @xmath145 for the expansive asw , we find the same functional forms for the solitary wave solutions in both cases . here , @xmath146 represents the compression of two adjacent particles i.e. , the strain field .
the solitary wave solution of eq . [ ross ] can be obtained by looking for propagating solutions of the form @xmath147 : @xmath148 which can be expressed in the form of newtonss - like equation @xmath149 = -\frac{dw}{d\delta}.\end{aligned}\ ] ] multiplying both sides by @xmath150 and integrating , @xmath151 integrating again , we find @xmath152 substituting , @xmath153 $ ] , and writing for brevity @xmath154 and @xmath155 , we need to integrate @xmath156 making the change of variables , @xmath157 , we find @xmath158 therefore , @xmath159 , \end{aligned}\ ] ] that yields , @xmath160 or @xmath161 squaring and substituting @xmath162 , @xmath163 that yields , @xmath164 therefore , the solitary wave solution is @xmath165
the energy and the fluctuations of the solitary wave can also be derived for a discrete chain , without making the continuum approximation . from eq .
( [ a_lagrangian ] ) , the equation of motion is @xmath166 a solitary wave solution to the above equation of motion has the form of a wave moving at a constant speed @xmath167 : @xmath168 where @xmath42 is the amplitude of the sw and @xmath169 is a function that describes the shape of the sw ( since we define the amplitude as the maximum speed of the particles enveloped by the solitary wave , the displacement @xmath170 will be proportial to @xmath171 . ) in analogy with eq .
( [ langevin_particle ] ) , we now couple the discrete equation of motion eq .
( [ discrete_eom ] ) to a source of gaussian noise and drag : @xmath172 here , the noise satisfies @xmath173 .
the coefficient of the noise ensures that in equilibrium the particles satisfy the fluctuation - dissipation theorem such that their average kinetic energy is @xmath174 . from eq .
( [ a_lagrangian ] ) , the energy of the chain is @xmath175 .
if we restrict the summation to particles enveloped by the solitary wave such that the sum does not include particles that are far away from the sw , then @xmath2 corresponds to the energy of the sw .
multiplying eq .
( [ discrete_noise ] ) by @xmath176 , we obtain the rate of change of energy @xmath177 where , the contribution of terms of the form @xmath178 ( i.e. , with index @xmath179 and @xmath180 ) cancel out upon summation over @xmath181 .
consistent with the quasi - particle interpretation of the sw , we again assume that the sw in a background of noise and drag , is still given approximately by eq .
( [ eq : train ] ) .
therefore , we write the terms in eq .
( [ eq : energyconservation ] ) in terms of the sw amplitude .
the sw energy is @xmath182 , and the first term on the right hand side is proportional to the kinetic energy , which is @xmath183 by the virial theorem .
thus the energy decays at a rate @xmath184 times the decay rate for the velocity of a single particle ( @xmath185 ) .
the sw amplitude therefore decays as : @xmath186 a term that compensates for the diffusion of @xmath42 is omitted here ( this term is derived in ito calculus ) , but it is only a small amount that is negligible if @xmath187 .
now , the different contributions in the last term just add up to a new noise term @xmath188 , which is also not correlated in time .
( assume that @xmath189 is the amplitude of the noise while @xmath190 has a noise with one unit of strength . ) the original noise term @xmath191 has correlations because @xmath192 depends on the noise at an earlier time .
but if the noise mostly just changes the amplitude of the sw by a random amount , and does not cause the particles in it to be displaced randomly relative to each other , then these correlations are not too important , and that is why they cancel out in the equation for @xmath42 .
the variance of the noise is therefore @xmath193 so @xmath194 .
this can be determined by the virial theorem since it is proportional to the kinetic energy of the sw , hence @xmath195 .
therefore the amplitude of the soliton satisfies @xmath196 upon taking the mean and variance of the amplitude from eq .
( [ discrete_a ] ) , we recover the solutions derived using continuum approximation . *
acknowledgments * nu acknowledges financial support from fom . v. f. nesterenko , _ j. appl . mech .
phys . _ * 5 * , 733 ( 1984 ) .
p. rosenau , _ phys .
a _ * 118 * , 5 ( 1986 ) . l. r. gmez , a. m. turner , m. van hecke and v. vitelli , _ phys
_ * 108 * , 058001 ( 2012 ) . v.
f. nesterenko , _ dynamics of heterogeneous materials _ ( springer - verlag , new york , 2001 ) . c. coste , e. falcon , and s. fauve , _ phys .
rev.e _ * 56 * , 6104 ( 1997 ) . c. daraio , v. f. nesterenko , e. b. herbold , and s. jin , _ phys .
e _ * 72 * , 016603 ( 2005 ) .
s. job , f. melo , a. sokolow , and s. sen , _ granular matter _ * 10 * , 13 ( 2007 )
. s. sen , j. hong , j. bang , e. avalos , and r. doney , _ physics reports _ * 462 * , 21 ( 2008 ) .
a. yethiraj , _ royal soceity of chemistry _ * 3 * , 1099 - 1115 ( 2007 ) .
chang , j.i .
cirac and h.j .
kimble , _ phys .
lett . _ * 110 * , 113606 ( 2013 ) .
z. cheng , j. shu , w.b .
russel and p.m. chaikin , _ phys .
lett . _ * 85 * , 7 ( 2000 ) .
zhirov , a.s .
pikovsky and d.l .
shepelyansky , _ phys .
e _ * 83 * , 016202 ( 2011 ) .
m. sheinman , c.p .
broedersz and f.c .
mackintosh , _ phys .
* 109 * , 238101 ( 2012 ) .
davydov , _ j. theor .
_ 66 , 379 ( 1977 )
. t. dauxois and m. peyrard , _ physics of solitons _ ( cambridge university press , 2006 ) .
c. s. ohern , l. e. silbert , a. j. liu , and s. r. nagel , _ phys . rev .
e _ * 68 * 011306 ( 2003 ) .
i. atsushi , l. berthier and g. biroli _ j. chem
. phys . _ * 138 * 12a507 ( 2013 ) n. xu , t. k. haxton , a. j. liu , and s. r. nagel , _ phys .
_ * 103 * 245701 ( 2009 ) .
we are using soliton " and solitary wave " interchangeably to refer to a localized lump of of energy ; we do not assume anything about the equations being integrable . for many waves or solitons , differentiating @xmath2 with respect to @xmath0 would give the group velocity or the speed of the wave s propagation , but this does not work for this case .
( @xmath0 is not the canonical momentum . )
m. manjunath , a.p .
awasthi , and p.h .
geubelle , _ phys
. rev.e _ * 85 * , 031308 ( 2012 ) .
n.g . van kampen , _
stochastic processes in physics and chemistry _
( elsevier science publishers , 2007 ) .
e. arevalo , f.g .
mertens , y. gaididei and a.r .
bishop , _ phys .
e _ * 67 * , 016610 ( 2003 ) .
gillespie , _
markov processes : an introduction for physical scientists _
( academic press , 1992 ) .
d.s . lemons , _ an introduction to stochastic processes in physics _ ( the john hopkins university press , 2002 ) .
allen and d.j .
tildsey , _
computer simulation of liquids _
( oxford , new york , 1987 ) . | we study a chain of anharmonic springs with tunable power law interactions as a minimal model to explore the propagation of strongly non - linear solitary wave excitations in a background of thermal fluctuations . by treating the solitary waves as quasi - particles , we derive an effective langevin equation and obtain their damping rate and thermal diffusion .
these analytical findings compare favorably against numerical results from a langevin dynamic simulation . in our chains composed of two sided non - linear springs ,
we report the existence of an expansion solitary wave ( anti - soliton ) in addition to the compressive solitary waves observed for non - cohesive macroscopic particles . in linear elastic solids , phonons are the basic mechanical excitations responsible for energy propagation . by contrast ,
an aggregate of macroscopic grains just in contact with their nearest neighbours constitute a novel elastic material where solitary waves or shocks replace phonons as the basic excitations @xcite .
the origin of these strongly non - linear waves can be traced to the fact that , unlike the case of harmonic springs , the repulsive force between two grains in contact does not depend linearly on the relative compression .
so far , little effort has been directed to determine the fate of these strongly non - linear excitations in a background of thermal fluctuations because temperature is clearly not a parameter relevant to the elastic response of macroscopic grains . however , granular aggregates at zero pressure are just one example of a broader class of materials that can be prepared in a unique mechanical state called _ sonic vacuum _ @xcite .
this term originally coined by nesterenko in the context of strongly non - linear granular chains designate a material characterized by a vanishing elastic moduli and linear speed of sound @xcite .
grafted colloidal particles @xcite and ultra - cold atoms in optical lattices @xcite are microscopic systems that allow for tunable non - linear interactions , while being naturally coupled to a source of fluctuation ( thermal or quantum ) .
these fluctuations restore rigidity and generate long wavelength phonon modes @xcite . however , the physics of very high amplitude strain propagation is still predominantly non - linear and resembles the state of sonic vacuum perturbed by background fluctuations , even if the interaction potentials are typically two sided , unlike granular ones .
the non - linear regime is particularly relevant for some biological systems , where energy transport occurs through localized non - linear excitations with energy significantly higher than the thermal energy @xcite .
moreover , systems such as polymer networks and colloidal glasses undergoing an unjamming transition are also characterized by vanishing elastic moduli as the coordination number or packing fraction are lowered towards the critical point @xcite .
the effect of thermal fluctuations on the _ non - linear _ response of materials undergoing an unjamming transition is relatively unexplored , despite they are obvious examples of a sonic vacuum state at zero temperature @xcite .
note , that in the case of jamming the linear elastic moduli can be lowered towards zero even if the microscopic interactions are harmonic , simply because there are not enough forces to prevent floppy motions . in this article
, we focus on strongly non - linear mechanical waves propagating in a background of small thermal fluctuations , a non - equilibrium problem that lies outside the realm of perturbation theory .
the starting point of conventional perturbation methods is a linear elastic solid , possibly at finite temperature , perturbed by small anharmonic terms .
by contrast , we adopt as a starting point the fully non - linear state of sonic vacuum whose elementary excitations are long - lived solitary waves @xcite .
subsequently we _ switch on _ temperature as a small perturbation that creates a background of thermal fluctuations . as a minimal model that is analytically tractable , we study impulse propagation in a one dimensional lattice of non - linear springs with a tune - able power law interaction . by coupling the lattice to a heat bath
, we then study the effects of the thermal fluctuations on the leading solitary wave generated in response to an impulse of energy much higher than the background thermal energy .
our approach in a nutshell is to treat the solitary wave as a quasi - particle and derive an effective langevin equation that describes its stochastic dynamics .
we corroborate our analytical predictions for the damping rate and thermal diffusion of the solitary waves with langevin dynamic simulations .
the sonic vacuum is usually studied with a chain of non - cohesive beads that only interact upon compression ( one - sided interaction ) .
the system with springs has some of the same properties as the sonic vacuum , since the sonic vacuum is a property of the non - linear power law interaction ( without a harmonic term ) .
in addition to the compressive solitary waves seen in a lattice of macroscopic grains with one - sided repulsive interaction , we report an accompanying anti - solitary wave solution for the lattice of non - linear springs with two sided interactions . )
momentum ( @xmath0 ) relation @xmath1 .
_ inset _ : zoom in of the leading sw - asw pair from left panel ( b).,scaledwidth=50.0% ] |
recently ( r@xmath1nyi ) entanglement entropy has become a center of wide interest in a broad array of theoretical physics such as string theory , condensed matter physics and quantum information theory .
it is useful tool when we investigate the distinctive features of various quantum states .
for example , it provides us with the method of classifying the quantum structure of several states in condensed matter physics , e.g. , @xcite .
the quantum entanglement is expected to be an important quantity which may shed light on the mechanism behind the ads / cft correspondence , e.g.,@xcite . in particular , the entanglement on the boundary is expected to be related to gravity in bulk e.g. , @xcite . therefore , it is important to reveal the fundamental properties of ( r@xmath1nyi ) entanglement entropy .
its remarkable feature has been found through researches @xcite over the past a few years .
a state is excited perturbatively so that the energy of its excitation is much smaller than the inverse of the subsystem s size .
if the size of a subsystem is small , its entanglement entropy is proportional to the energy included in it . thus its entanglement entropy obeys the law analogous to the first law of thermodynamics . in relativistic setup ,
its effective temperature is proportional to the inverse of the subsystem s size . and
that temperature does not depend on the parameters of strongly coupled gauge theories in large @xmath2 limit .
in other words , it is universal in strongly coupled gauge theories in this limit . in those works , we studied the property of the entanglement entropy when the size of the subsystem is very small . in this paper and previous one @xcite , we study the property of ( r@xmath1nyi ) entanglement entropy when its size is infinite .
moreover we define the excited state by acting various local operators on the ground state .
( r@xmath1nyi ) entanglement entropies for them show nontrivial time evolution .
their excess finally approach certain finite values . before we explain how to compute them and explain the results which we obtain
, we would like to explain quantum quenches briefly .
there are two classes of quenches ; global quenches and local quenches .
the global quenches are triggered by changing parameters homogeneously , e.g. , @xcite . when parameters are changed homogeneously , we are able to study the thermalization of a subsystem and its thermalization time by measuring its entanglement entropy .
their holographic duals are the formation of a black hole , e.g. , @xcite .
the local quenches are triggered by adding an interaction locally to a hamiltonian or changing parameters locally in the hamiltonian @xcite .
when parameters are changed inhomogeneously , we can investigate the time evolution of the entanglement entropy locally .
a falling particle in ads spacetime is proposed as the holographic duals of these quenches , e.g. , @xcite . in this paper
, we introduce a new class of excited states .
these states are defined by acting a local operator on the ground state .
this local operator violates the time translational invariance and excites the ground state . in this setup
, we can study the time evolution of ( r@xmath1nyi ) entanglement entropies for these states .
therefore we can investigate how the entanglement propagates locally .
we would like to briefly explain how to compute the ( r@xmath1nyi ) entanglement entropies for these states and give an explanation of the results we obtain . in this work ,
we compute them in the free massless scalar field theory by the replica method . here
we choose the subsystem a to be a half of the total space .
we define a locally excited state by acting a local operator @xmath3 on the ground state .
we compute the ( r@xmath1nyi ) entanglement entropies for these states by using the replica method . after computing them ,
we perform an analytic continuation to real time and study their time evolution .
we find they finally approach certain finite constants .
the time evolution of the ( r@xmath1nyi ) entanglement entropies for locally excited states can be interpreted in terms of the relativistic propagation of entangled pairs . in previous paper @xcite , we derived the ( r@xmath1nyi ) entanglement entropies at late time for the state which is excited by @xmath3 from the entangled pair interpretation .
we call the final values of ( r@xmath1nyi ) entanglement entropies ( _ r_@xmath1_nyi _ ) _ entanglement entropies of operators_. they include the ( r@xmath1nyi ) entanglement entropy for an epr state .
they measure the degrees of freedom of local operators and characterize them in conformal field theories in the viewpoint of quantum entanglement .
finally we would like to summarize the new results we obtain in the present work . in previous paper @xcite
, we found that the ( r@xmath1nyi ) entanglement entropies of @xmath3 is given by those of binomial distribution only when the replica number and @xmath4 is small .
in present paper , we provide the explanation which shows that the ( r@xmath1nyi ) entanglement entropies of @xmath3 are given by those of binomial distribution by using the replica method .
the results which we obtain by using replica method agree with those which we obtain in terms of the entangled pair .
we also obtain the sum rule for the ( r@xmath1nyi ) entanglement entropies for the states defined by acting the various operators on the ground state .
they are given by the sum of the ( r@xmath1nyi ) entanglement entropies for the states defined by acting one of these operators on the ground state .
moreover we argue that they can be generalized .
we limit operators to the operators @xmath5 which are composed of single - species operator .
the ( r@xmath1nyi ) entanglement entropies of them are given by those of binomial distribution .
the entropies of general operators obey the same sum rule as the one which we mentioned above .
this paper is organized as follows . in section @xmath6 , we compute the ( r@xmath1nyi ) entanglement entropies for the states defined by acting various local operators on the ground state in the replica method .
we also compute the green function on the space which has a conical singularity .
we explain an analytic continuation which is useful to investigate the time evolution of those entropies for excited states . in section @xmath7 , we calculate ( r@xmath1nyi ) entanglement entropies for the states which excited by scalar operators and investigate their time evolution explicitly . in section @xmath8
, we interpret their time evolution in terms of entangled pair .
we find that the ( r@xmath1nyi ) entanglement entropies of @xmath3 is given by those of binomial distribution . in section @xmath9
, we explain the new results we obtain in this paper .
we provide the explanation which shows that the ( r@xmath1nyi ) entanglement entropies of @xmath3 are given by those of binomial distribution by the replica method .
we argue that those results can be generalized . in section @xmath10
, we summarize the conclusions and discuss the future problems .
let us study the time evolution of ( r@xmath1nyi ) entanglement entropies for the states generated by acting various local operators on the ground state . to investigate their time evolution
, we will compute them by the replica methods .
after computing the ( r@xmath1nyi ) entanglement entropies for those states by the replica method , we will perform an analytic continuation to real time and investigate their real time evolution . in this section
, we will explain the analytic continuation which is used to investigate their time evolution and also explain how to compute them .
those entropies are related to the correlation functions of these operators as we explain later . in this paper
, we compute them in free massless scalar field theory in even dimension .
therefore we compute the green function on @xmath11-sheeted geometry in even dimensional free massless scalar field theory . in this subsection
, we would like to explain the analytic continuation which is used to investigate the real time evolution of the ( r@xmath1nyi ) entanglement entropies for the states defined by acting various operators on the ground state .
we consider a qft in @xmath12 dimensional spacetime @xmath13
let us define the locally excited state by @xmath15 where @xmath16 is located at @xmath17 ( @xmath18 ) .
its density matrix is given by @xmath19 where @xmath20 is determined so as to @xmath21 .
this local operator is acted on the ground state at @xmath22 and triggers the time evolution of quantum entanglement in the total system .
we investigate the time evolution of quantum entanglement by measuring ( r@xmath1nyi ) entanglement entropy at @xmath23 . to compute the ( r@xmath1nyi ) entanglement entropy for these locally excited states , we firstly introduced the regularization parameter @xmath24 . by using this parameter ,
@xmath25 is deformed as follow , @xmath26 high energy modes are suppressed by @xmath27 .
thus its ( r@xmath1nyi ) entanglement entropies are regulated . to compute the ( r@xmath1nyi ) entanglement entropies for this state
, we use the replica methods . to compute them by the replica method
, we define a density matrix by @xmath28 where @xmath29 , @xmath30 respectively .
@xmath24 is the regularization parameter for two point function of @xmath16 and @xmath31 .
after computing ( r@xmath1nyi ) entanglement entropies for this state by replica method , we perform the analytic continuation to real time as follows , @xmath32 after analytic continuation , the relation between the polar coordinate and @xmath33 is shown in fig.1 .
thus we investigate the evolution of @xmath34 for excited states given by ( [ exm ] ) with @xmath35 .
[ pro2 ] and @xmath36 are related to @xmath33 and @xmath37 as in this figure.,title="fig:",width=302 ] let us compute ( r@xmath1nyi ) entanglement entropies for the density matrix defiend by ( [ rhoe2 ] ) .
the local operators @xmath16 and @xmath31 are located at @xmath38 and @xmath39 respectively . to define ( r@xmath1nyi ) entanglement entropies , we choose subsystem a to be a half space @xmath40 .
tracing out the degrees of freedom in subsystem b , we define a reduced density matrix by @xmath41 the n - th r@xmath1nyi entanglement entropy for this state is defined by @xmath42 we would like to study the excess of ( r@xmath1nyi ) entanglement entropies for this state defined by ( @xmath43 ) .
therefore we subtract those entropies for the ground state from the entropies for the locally excited state and we define their excess by @xmath44 @xmath45 is the normalized reduced density matrix for the ground state .
let us compute @xmath34 by path - integral as in @xcite .
it is useful to introduce a polar coordinate .
the location of operators is shown in fig.1 .
then the @xmath55 is given by the partition function where @xmath56 operators are inserted , @xmath57 } , \end{split}\ ] ] where operators are periodically located on the @xmath58sheeted geometry @xmath59 and @xmath60 .
@xmath59 is constructed gluing the subsystem a on a sheet to subsystem a on the next sheet as in fig.2 .
[ rep ] -sheeted geometry @xmath59 is constructed by gluing subsystem a on a sheet to subsystem a on the next sheet .
fields acquire the phase shift of @xmath61 if they go around the boundary of a on the each sheet of @xmath59.,title="fig:",width=302 ] and the reduced density matrix for the ground state is given by @xmath62 where @xmath63 is the partition function on the @xmath58sheeted geometry and @xmath64 is the partition function on @xmath65 . substituting ( [ ep ] ) and ( [ gp ] ) into ( [ sre ] ) , the path integral representation of @xmath66 is given by @xmath67 the first term of the last line in ( [ gre ] ) is given by a @xmath56 point function of @xmath16 and @xmath31 on @xmath59 . and the second term in ( [ gre ] ) is given by @xmath6 point function of @xmath16 and @xmath31 on @xmath68 . here @xmath68 is usual @xmath12 dimensional euclidean space .
here only one operator is acted on the ground state .
however the result in ( [ gre ] ) can be generalized to that for the excited state defined by acting several operators on ground state as we will explain later . in previous subsection , we defined ( r@xmath1nyi ) entanglement entropy for locally excited states in general field theories . from now on
, we will explain how to compute those entropies and their evolution with time in free massless scalar field theory .
@xmath34 are related to the @xmath56 point function on @xmath59 and @xmath6 point function on @xmath68 as in ( [ gre ] ) .
therefore we explain how to compute the green function of @xmath69 on even dimensional @xmath59 .
we choose the region of subsystem a to be a half @xmath70 of the total space as in fig.3 .
we introduce the polar coordinate as @xmath71 @xmath72 .
operators are located at @xmath73 respectively .
@xmath74 are the coordinates along the direction vertical to the plane which has a conical singularity.,title="fig:",width=302 ] the green function on @xmath59 is defined by @xmath75 where @xmath76 and @xmath77 .
then we can expand the green function in terms of the eingenfunctions @xmath78 , @xmath79 where @xmath80 .
as in @xcite , the green function on @xmath59 is given by @xmath81 where @xmath82 and @xmath83 is the bessel function of the first kind . to integrate with respect to @xmath84 in right hand side of the equation in ( [ 1 ] ) , we express it in terms of a schwinger parameter . after that
, its denominator is given by @xmath85 after express it in terms of schwinger parameter , we integrate with respect to @xmath84 in the right hand side of that equation , @xmath86 moreover , we integrate with respect to @xmath4 .
if we use the formula of bessel function , the integral of @xmath4 from @xmath87 to @xmath88 in ( [ 1 ] ) can be performed , @xmath89 where @xmath90 is the modified bessel function of first kind .
@xmath90 is given by @xmath91 where the contour in the complex plane is shown in fig.4 .
[ ccp ] after performing integral of @xmath92 , @xmath93 is given by @xmath94 when d is odd , we can perform integration of t. below we will mainly investigate the time evolution of ( r@xmath1nyi ) entanglement entropy for excited states in @xmath8 and @xmath10 dimensional cases . to evaluate ( [ gre ] ) , we would like to give the explicit form of the green function on @xmath8 and @xmath10 dimensional @xmath59 . when @xmath95 is @xmath7 , @xmath96 is given by @xmath97 where the parameter @xmath98 is given by @xmath99 as in @xcite . when @xmath95 is @xmath9 , @xmath93 is given by @xmath100 .
\end{split}\ ] ] by using these green functions , we will study the time evolution of ( r@xmath1nyi ) entanglement entropies for locally excited states .
let us study the time evolution of ( r@xmath1nyi ) entanglement entropies for various locally excited states in @xmath101 and @xmath10 dimensional free massless scalar field theory .
they will approach a finite constant value at late time . here
, we will explain how to compute them .
first , we will study them in two dimensional examples . in these examples
, we can compute the r@xmath1nyi entanglement entropies for locally excited states by using a conformal map .
we will explain the results of @xmath102 for the states defined by acting several operators on the ground state .
we choose @xmath103 , @xmath104 , @xmath105 , @xmath106 and @xmath107 as the operators ( @xmath108 is a complex number ) .
we define the excited state by acting them on the ground state .
@xmath102 for the last one shows nontrivial time evolution .
we assume these operators are located at @xmath109 . in 4 and 6 dimensions ,
we define the excited states by acting @xmath3 on the ground state .
@xmath34 for those states show the similar time evolution and approach the finite constants at late time in both dimensions .
@xmath34 in @xmath10 dimension increase milder than those in @xmath8 dimension .
however the final values of @xmath34 do not depend on spacetime dimension but depend on @xmath4 and @xmath11 .
therefore we explain would like to explain how to compute them in only @xmath8 dimension . here
we compute them in two illustrative examples . in the first example
, we compute @xmath102 for the state which is excited by a the local operator @xmath69 .
we also compute that for the state which is excited by the two local operators @xmath110 in the second example . in the latter example
, we assume the two operators are located separately . in the latter example , we would like to study the time evolution of @xmath102 in the two setups . in first setup , one operator @xmath69
is located at @xmath111 and another one is located at @xmath112 where @xmath113 . in the region
@xmath114 , @xmath102 is given by a finite constant which depends on the parameters @xmath115 .
in the region @xmath116 , @xmath102 shows nontrivial time evolution and finally approaches a certain constant value .
it also depends on only two parameters @xmath115 . in both cases ,
@xmath102 is given by the sum of the ( r@xmath1nyi ) entanglement entropy for the state which is excited by @xmath69 .
in this subsection , we explain how to compute @xmath34 for locally excited states and also explain the results in two dimensional examples .
we compute @xmath34 for the states defined by acting the derivatives of @xmath69 and exponentials of @xmath69 on the ground state .
we choose @xmath103 , @xmath117 and @xmath118 as derivatives of @xmath69 . @xmath106 and @xmath107 are chosen as its exponentials . in two
dimensional free massless scalar filed theory , we can calculate @xmath34 by using a conformal map @xmath119 . in this case
, we do not need the green function which we obtain in section 2 . here
@xmath120 is the coordinate of @xmath11-sheeted geometry and @xmath121 is the coordinate of the flat plane . to compute derivatives and exponentials of @xmath69
, we just have to obtain the two point function of @xmath69 . the green function on the flat plane @xmath122
is given by @xmath123 we impose that the behavior of the two point function on @xmath59 is the same as the behavior of the green function on @xmath68 if we take the @xmath124 limit .
we also impose that this two point function on @xmath59 satisfies the properties of green function . by using the conformal map
@xmath125 , two point function of @xmath69 on @xmath59 is given by @xmath126 after substituting the propagators in ( [ epro ] ) into ( [ gre ] ) , we perform the analytic continuation defined in ( [ ac ] ) to real time . after that , we can study time evolution of @xmath127 for the states defined by acting local operators on the ground state . here
we compute the @xmath102 for locally excited states defined by acting the derivatives of @xmath69 on the ground state .
these operators are limited to @xmath128 , @xmath129 , @xmath130 .
after computing the second r@xmath1nyi entanglement entropies for these states by the replica method , we perform the analytic continuation in ( [ ac ] ) and take the @xmath131 limit . in this limit ,
they vanish , @xmath132 as we explain in the subsection 4.2 , these results can be physically interpreted in terms of relativistic propagation of quasi - particles .
we would like to explain the results of the excited states generated by acting exponential operators of @xmath69 on the ground state such as @xmath106 , @xmath133 .
@xmath102 for the state which is excited by @xmath134 vanishes .
on the other hand , @xmath102 for the state defined by acting @xmath133 on the ground state does not vanish . in the replica method ,
@xmath102 for this state is given by @xmath135},\ ] ] where @xmath16 and @xmath31 are given by @xmath136 by substituting propagators in ( [ epro ] ) into ( [ 2se ] ) , we perform the analytic continuation which is defined in ( [ ac ] ) and take the @xmath131 limit .
the time evolution of @xmath127 is given by the step function , @xmath137 } & t_1 > 0 .
\\ \end{cases}\ ] ] @xmath102 does not depend on the conformal dimension of operator @xmath16 .
this final value depends on only the complex constant @xmath108 .
when @xmath138 is @xmath139 , the final value of @xmath102 is given by its maximum value .
it is equivalent to the second r@xmath1nyi entanglement entropy for an epr state .
the time evolution of @xmath102 is plotted in fig.5 . in subsection 4.1
, we interpret this result in terms of entangled quanta .
[ 2dex ] as functions of @xmath35 in the @xmath131 limit .
the vertical line corresponds to @xmath102 .
and the horizontal line corresponds to @xmath35 .
here we chose @xmath140 .
the red curve corresponds to the plot of @xmath102 for @xmath141 .
the blue curve corresponds to the plot of @xmath102 for @xmath142.,title="fig:",width=302 ] in this subsection , we compute @xmath102 and study its time evolution in two examples . in both cases , we compute @xmath102 for locally excited states by the replica method .
after computing it , we perform the analytic continuation defined by ( [ ac ] ) to real time and we study its real time evolution .
as we mentioned above , we compute @xmath102 for the excited state defined by acting a local oprerator @xmath69 in @xmath8 dimensional free massless scalar field theory .
let us calculate its second r@xmath1nyi entanglement entropy and study its time evolution .
we assume that @xmath69 is located at @xmath143 .
this locally excited state is given by @xmath144 to investigate the time evolution of @xmath102 for this locally excited state , we computed @xmath102 by the replica method .
this operator is located at @xmath145 .
this state is given by @xmath146 after mapping @xmath147 to @xmath148 as in fig.6 , @xmath102 is given by @xmath149},\ ] ] where @xmath150 is located at @xmath151 as in fig.6 .
[ smap ] to compute @xmath102 , we need to compute @xmath8 point correlation function of @xmath69 on @xmath152 and @xmath6 point correlation function of @xmath69 on @xmath68 .
therefore we compute the green function on @xmath152 and @xmath68 . and
these green functions are respectively given by @xmath153 by substituting these propagators in ( [ 4pro ] ) into ( [ phi1 ] ) , we perform the analytic continuation defined by ( [ ac ] ) to real time . in this analytic continuation ,
the following useful relations hold @xmath154 after performing this analytic continuation , we take the @xmath131 limit . in this limit , the two point function on @xmath155 is given by @xmath156 and the leading term of four point function on @xmath152 is @xmath157 in this limit .
only specific propagators can contribute to this four point function .
we call these propagators _ dominant propagators_. in the region @xmath158 , dominant propagators are given by @xmath159 in the region @xmath160 , dominant propagators are given @xmath161 the other propagators are at most @xmath162 in both regions .
therefore we can ignore them in the @xmath131 limit . in the region @xmath163 , the four point correlation function of @xmath69 on @xmath152
is given by @xmath164 in the region @xmath160 , that is given by @xmath165 therefore , the second r@xmath1nyi entanglement entropy for the excited state defined by acting @xmath69 on the ground state is given by @xmath166}.\ ] ] the time evolution of @xmath102 in ( [ sep1 ] ) is plotted in fig.7 .
[ sl10 ] with @xmath35 in the @xmath131 limit .
the vertical axis corresponds to @xmath102 .
the horizontal axis corresponds to @xmath35 .
here we chose @xmath140 .
the blue curve is the @xmath102 for @xmath167 .
the red curve is the final value of @xmath102 .
it is given by @xmath168 .
, title="fig:",width=302 ] if we take the @xmath169 limit , its final value is given by @xmath170 let us compute the second r@xmath1nyi entanglement entropy for the state defined by acting two local operators @xmath110 on the ground state .
we assume these operators are located separately .
we compute it in two different examples . in first example
, we act @xmath69 on the ground state at time @xmath171 and act another one on it at time @xmath172 .
we investigate evolution of the second r@xmath1nyi entanglement entropy with time @xmath173 . in second example
, we act two operators on the ground state simultaneously .
we also investigate the time evolution of the second r@xmath1nyi entanglement entropy .
let us compute @xmath102 for the state which is excited by two local operators @xmath110 .
one of those operators is located at @xmath174 as in fig.8 . here
, we assume that @xmath175 .
if @xmath176 , then the time evolution of @xmath102 is not affected by local operator @xmath177 .
another one is located at @xmath178 .
we would like to investigate the evolution of @xmath102 with time @xmath179 .
[ mana ] is related to @xmath180 and @xmath24 after performing an analytic continuation to real time.,title="fig:",width=302 ] in this example , the excited state is given by @xmath181 where @xmath182 is the time ordering operator . in the replica method , @xmath102 for this state is given by @xmath183 } \end{split}\ ] ] to compute @xmath102 , we have to compute the eight point correlation function of @xmath69 on @xmath152 and the four point function on @xmath68 .
therefore we need to compute green function on @xmath152 and @xmath68 .
these green functions are given by ( [ 4pro ] ) .
in the replica method , @xmath184 corresponds to @xmath185 and @xmath186 and they are located at @xmath187 and @xmath188 respectively . on the other hand , @xmath189 corresponds to @xmath190 , @xmath191 in that method and they are located at @xmath192 and @xmath193 respectively as in fig.8 .
in addition to the correlations in the case where one operator is acted on the ground state , there are the correlations which correspond to the correlation between @xmath194 and @xmath195 such as @xmath196 , @xmath197 @xmath198 .
it seems that we have to take these correlation functions into account .
however these correlation functions can not contribute to @xmath102 when we take the @xmath131 limit .
therefore we can neglect the contribution from them .
below we will explain why these correlation functions can not contribute to @xmath102 in this limit .
after substituting the propagators in ( [ 4pro ] ) into ( [ dtwo ] ) , we perform an analytic continuation as follow , @xmath199 as in fig.8 . in this continuation , there are useful relations as follows , @xmath200 and @xmath201 after performing this analytic continuation , we take the limit @xmath131 . in this limit ,
4 point function on @xmath68 is given by @xmath202 on the other hand , the leading term of 8 point function of @xmath69 on @xmath152 is @xmath203 in this limit . in the region @xmath204 , the leading term of 8-point function of @xmath69
is given by @xmath205 where @xmath206 denotes the @xmath207 point function .
in the region @xmath114 , that of 8 point function of @xmath69 is given by @xmath208 in this limit , the leading term of 8-point function comes from the contributions from only spesific propagators .
their leading term is @xmath209 . in the region
@xmath204 , @xmath210 , @xmath211 , @xmath212 , @xmath213 , @xmath214 and @xmath215 can contribute to the leading term of 8 point function . in this limit , these propagators are given @xmath216 in the region @xmath116 , the contributions from @xmath212 , @xmath213 , @xmath214 and @xmath215 do not change . on the other hand , the contributions from @xmath217 , @xmath218 change as follows , @xmath219 @xmath220 and @xmath218 can also contribute to leading term of 8-point function of @xmath69 .
these propagator are given @xmath221 the contribution from other propagators is at most @xmath222 .
then they can not contribute to the leading term of 8 ponit function of @xmath69 . in the @xmath131 limit ,
@xmath102 for this state is given by @xmath223 } & t_2\le l , \\
\log{\left[\frac{4 t_1 ^ 2 t_2 ^ 2}{\left(l^2+t_2 ^ 2\right ) \left(l^2+t_1 ^ 2\right)}\right ] } & t_2 \ge l. \end{cases}\ ] ] [ ml10 ] with @xmath179 in the @xmath131 limit . the vertical axis corresponds to @xmath102 .
the horizontal axis corresponds to @xmath179 .
here we chose @xmath140 .
the blue curve corresponds to the time evolution of @xmath102 for @xmath224 , where one of two scalar fields is located at @xmath225 .
the green curve corresponds to the time evolution of @xmath102 for @xmath224 where one of two scalar fields is located at @xmath226 .
the red curve corresponds to @xmath227.,title="fig:",width=302 ] its time evolution is plotted in fig.10 .
the final value of @xmath102 is given by @xmath228}\ ] ] we are able to control its final value by turning the parameters @xmath35 and @xmath229 . and the maximum value of it is given by @xmath230 let us think about the case where we act two operators @xmath110 simultaneously on the ground state .
these operators are located at @xmath231 and @xmath232 respectively . here
, we assume that @xmath233 .
we investigate the evolution of the second r@xmath1nyi entropy with @xmath234 for this state .
we can compute it as we explained above .
we just have to change @xmath179 to @xmath35 in ( [ 2ana ] ) .
therefore we skip the detail of the calculation . of course
, we perform the analytic continuation similar to that in ( [ 2ana ] ) and take the @xmath131 limit .
the time evolution of second r@xmath1nyi entanglement entropy is given by @xmath235 } & t_1 \ge l , \\
\log{\left[\frac{2t_1 ^ 2}{l^2+t_1 ^ 2}\right]}+ \log{\left[\frac{2t_1 ^ 2}{l^2+t_1 ^ 2}\right ] } & t_1
\ge l. \end{cases}\ ] ] the evolution of @xmath102 with @xmath35 is plotted in fig.11 . if we define the entropy for the state which is excited by @xmath177 and @xmath236 by @xmath237 and @xmath238 respectively , then the entropy for the state generated by acting both of these operators on the ground state is given by the sum of these entropies , @xmath239 if we take the @xmath240 limit , the final value of @xmath102 is given by @xmath241 [ l10l40 ] with @xmath35 in the @xmath131 limit .
the vertical axis corresponds to @xmath102 .
the horizontal axis corresponds to @xmath35 .
here we chose @xmath242 and @xmath243 .
the blue curve is the time evolution of @xmath102 for @xmath224 where one of the two scalar fields are spatially located at @xmath244 .
another one is located at @xmath245 the purple curve is @xmath168 .
the red curve is @xmath227 .
, title="fig:",width=302 ]
in previous section , we investigate the time evolution of @xmath34 . here
we will interpret the time evolution of @xmath34 in terms of relativistic propagation of entangled quanta .
moreover we will acquire @xmath246 for the state defined by acting the local operator @xmath3 on the ground state by using this framework .
we will also explain the new results which we obtain in present paper .
one of them is the large @xmath4 behavior of @xmath246 for the state defined by acting the local operator @xmath3 on the ground state .
we will find the sum rule for the ( r@xmath1nyi ) entanglement entropies for the state defined by acting several operators on the ground state .
we assume that these operators are given by @xmath247 and located separately .
we will argue @xmath246 for the state defined by acting specific operators on the ground state is given by ( [ gfo ] ) .
we will also argue that the sum rule can be generalized .
we investigated the time evolution of @xmath34 for the locally excited states in 2,4 and 6 dimensional spacetime in previous section . here
we interpret the time evolutions of them in terms of relativistic propagation of the entangled pair .
let us consider the result of @xmath248 in two dimension in terms of its propagation .
the time evolution of @xmath102 for this state corresponds to the blue curve in fig.5 . in this figure
, the amount of @xmath102 can not increase until @xmath249 .
it drastically increases at @xmath249 and reaches constant value @xmath168 at late time .
we can interpret its time evolution in terms of relativistic propagation of entangled quanta as follows .
first , a pair of two quanta appears at the point where that operator is located as in fig.11 . here
this operator is assumed to be inserted into the complement b of subsystem a. these two quanta are entangled with each other and we call this pair an entangled pair . each of two quanta propagates in the opposite direction at the speed of light . until @xmath249 , two quanta remain to stay in the region b. at that time , the quantum entanglement between them is not able to contribute to @xmath102 .
when one of them reaches at the boundary @xmath250 of subsystem a , the quantum entanglement between them begins to contribute to the second r@xmath1nyi entanglement entropy between a and b. since these quanta propagate at the speed of light , one of them can reach @xmath250 at @xmath249 .
we chose a half of total space as the subsystem a. therefore , once one of them enters the subsystem a , it can not go out there . since each of entangled pair remain to stay in a and b respectively , @xmath127 can approach a finite constant . in particular , these quanta can propagate in only one direction respective in two dimensional examples .
therefore @xmath102 suddenly reaches a nontrivial constant at @xmath249 . in higher dimensional examples(@xmath251 ) , those quanta can propagate in generic direction .
then @xmath34 approaches some constants values mildly as in fig.7 . .
at @xmath158 , each of entangled pair is included in region b. at @xmath252 , two quanta are included in a and b respectively .
, width=302 ] . at @xmath158 ,
each of entangled pair is included in region b. at @xmath252 , two quanta are included in a and b respectively .
, width=302 ] moreover we are able to find out why @xmath253 is given by @xmath168 by using this interpretation
. we decompose @xmath69 into @xmath254 and @xmath255 , @xmath256 where @xmath254 and @xmath255 respectively correspond to the modes which are moving towards the left direction @xmath257 and the right direction @xmath258 .
therefore this locally exited state is given by @xmath259 where @xmath260 is normalized so as to @xmath261 .
this locally excited state is an epr state ( a maximally entangled state )
. therefore @xmath253 is given by @xmath168 .
for any @xmath11 , we can acquire @xmath34 for that state .
its reduced matrix is given by @xmath262 and for any @xmath11 , @xmath263 is given by @xmath264 the definition of ( r@xmath1nyi ) entanglement entropies are given in ( [ drenyi ] ) .
( r@xmath1nyi ) entanglement entropies for this rpr state do not depend @xmath11 .
therefore they are given by @xmath265 let us apply this interpretation for other results in two dimensional cases . in the case of the entropy for the excited state defined by acting the derivatives of @xmath69 mentioned above and @xmath266 on the ground state , these locally excited states can be represented in terms of products states . therefore @xmath34 vanishes . in next subsection , we apply the entangled pair interpretation for the results in higher dimensional cases .
let us apply the entangled pair representation for the results in higher dimensional cases .
and we will obtain @xmath246 for the states defined by acting @xmath3 on the ground state .
we decompose @xmath69 into @xmath254 and @xmath255 as in @xmath268 . using this decomposition
, the excited state defined by acting @xmath3 on the ground state is given by @xmath269 where @xmath20 is the normalization factor and@xmath270 and @xmath271 are the vacuum for the right moving mode and the left moving mode respectively .
moreover we define new bases by , @xmath272 where these states are normalized so as to @xmath273 .
and @xmath274 is also normalized so as to @xmath275 .
after normalized , this locally excited state is given by @xmath276 if we take @xmath240 limit , we can identify the degrees of freedom of right moving modes and left moving modes with those in subsystem a and b respectively . therefore tracing out the degrees of freedom in subsysten b
is equivalent to tracing out those of left moving modes .
after tracing out the degrees of freedom in subsystem b , its reduced density matrix is give by @xmath277 where @xmath278 .
it is given by binominal distribution . by using the matrix in ( [ gmat ] )
, we can obtain its explicit form for each k as follows , @xmath279 substituting @xmath280 in ( [ gmat ] ) into ( [ drenyi ] ) , we are able to acquire the @xmath11-th r@xmath1nyi entanglement entropy for the locally excited state defined by acting @xmath3 on the ground state , @xmath281 it is given by the @xmath11- th r@xmath1nyi entanglement entropy of binomial distribution
. we can derive the explicit results for each @xmath282 from the formula in ( [ gfo ] ) .
these results do not depend on spacetime dimension as long as its dimension is higher than @xmath6 . as you see later , they are consistent with results which we will obtain by the replica method .
these results are summarized in table.1 . in any dimensions ,
the final values of the ( r@xmath1nyi ) entanglement entropies for the state which is excited by @xmath3 are given by ( [ gfo ] ) .
in @xmath283 case , the formula in ( [ gfo ] ) is simplified , @xmath284 @xmath285 does not depend on the number of sheets .
it is equivalent to the ( r@xmath1nyi ) entanglement entropy for an epr states ( maximally excited state ) . using the formula in ( [ gfo ] ) ,
we can compute @xmath286 for the state defined by acting @xmath3 on the ground state . as we will see later
, the formula in ( [ gfo ] ) gives a characterization of operators .
we define the ( r@xmath1nyi ) entanglement entropies of operators by the final value of ( r@xmath1nyi ) entanglement entropies for the state defined by acting them on the ground state .
it characterizes the quantum entanglement of operators and it is independent of the conformal dimension of operators .
the quantum entanglement of operators is related to the number of entangled pair which those operators are able to create at least in free massless scalar field theory .
.the value of @xmath287 for free massless scalar field theories in dimensions higher than two .
here @xmath11 is the number of sheet .
@xmath288 is the number of @xmath69 [ cols="^,^,^,^,^,^,^",options="header " , ] here we explain one of the new results .
we would like to comment on the large @xmath4 behavior of @xmath34 .
the ( r@xmath1nyi ) entanglement entropies of @xmath3 are given by those of binomial distribution .
therefore in large @xmath4 limit we can approximate its reduced density matrix by normal distribution as follows @xmath289 where @xmath98 is a certain positive constant . by using this approximation , we would like to evaluate @xmath290 in table.1 .
@xmath290 are given by @xmath291 they do not depend on the number of sheets .
we explained the ( r@xmath1nyi ) entanglement entropies of @xmath3 up to here .
let us consider the ( r@xmath1nyi ) entanglement entropy for the states defined by acting specific operators on the ground state here .
we limit operators to particular operators such as @xmath293 .
they are composed of single - species operators such as @xmath69 , @xmath103 , @xmath294 and so on .
and we call these operators s - operators . in the argument above
, we assumed that operators can be decomposed to left moving modes and right moving modes at late time . in this assumption , we did not use the detailed feature of local operator @xmath69 .
therefore , we argue that we are able to decompose these operators @xmath295 , @xmath103 , @xmath296 into the right moving mode and left moving mode when states are excited by s - operators . therefore in free massless scalar field theory in any dimension ,
the ( r@xmath1nyi ) entanglement entropies of s - operators are given by the formula in ( [ gfo ] ) .
in summary , we argue that the ( r@xmath1nyi ) entanglement entropies of s - operators should be given by ( [ gfo ] ) .
this implies that quantum entanglement of operators is independent of the conformal dimensions of operators .
the ( r@xmath1nyi ) entanglement entropies of operators characterize operators in the viewpoint of quantum entanglement .
we have explicitly computed the quantum entanglement of @xmath297 , for @xmath298 in @xmath299 .
we found that they are consistent with the results we obtained under this assumption .
moreover , @xmath300 which we derive from the entangled pair interpretation agrees with the result which we will obtain by the replica method
. we will interpret the final value of ( r@xmath1nyi ) entanglement entropy for states by acting various operators on the ground state in terms of entangled pair .
first , we think about @xmath34 for @xmath301 . these operators are located separately .
we assume @xmath302 , @xmath303 can be independently decomposed to left moving modes and right moving modes as follow , @xmath304 where @xmath305 and @xmath306 are independent of other operators @xmath307 .
then the annihilation and creation operators of @xmath308 can commute with those of @xmath309 . under this assumption ,
the normalized excited state is given by @xmath310 where @xmath182 is the time ordering operator .
in any dimensions , the final value of the @xmath11-th r@xmath1nyi entanglement entropy for this state is given by @xmath311 where @xmath312 and @xmath313 are given by ( [ gfo ] ) .
they are the ( r@xmath1nyi ) entanglement entropy of @xmath314 and @xmath315 respectively .
@xmath300 is given by the sum of @xmath312 and @xmath313 .
when @xmath316 , @xmath317 are @xmath139 respectively in @xmath283 case , @xmath318 is consistent with the result which we obtained in previous section .
the sum rule in ( [ 2sum ] ) can be generalized .
we prepare the locally excited state defined by acting several operators on the ground state .
these operators @xmath319 @xmath320 are given by @xmath247 .
we assume they are located separately .
in any dimensions @xmath267 , @xmath300 at late time for the state generated by acting these operators on the ground state are given by @xmath321 where @xmath322 is the ( r@xmath1nyi ) entanglement entropy of the local operator @xmath319 .
as you see later , the result in ( [ gsum ] ) agrees with the result which we obtain by replica method .
we would like to generalize the sum rule of ( r@xmath1nyi ) entanglement entropies of operators to those in more general case where operators are not limited to @xmath247 .
we did not use the distinctive feature of @xmath69 when we found the sum rule of @xmath300 .
we assumed that @xmath69 can be decomposed to left moving mode and right moving mode .
if the location of @xmath323 is separated from the location of @xmath324 , @xmath323 is independent of @xmath324 .
it is expected that we can apply this assumption for the ( r@xmath1nyi ) entanglement entropy of operators other than @xmath3 .
therefore we argue that if the state is defined by acting various general operators @xmath325 on the ground state , @xmath300 for that state should obey the sum rule which is given by ( [ gsum ] ) . here
, we assume that @xmath325 are located separately .
in other words , @xmath300 for that state is given by the sum of the ( r@xmath1nyi ) entanglement entropies of those operators in any dimensions .
in summary , the excited state is given by @xmath326 where the locations of operators are separated from the locations of operators other than itself .
we assume that @xmath319 are general operators . in any dimensions @xmath267 , the ( r@xmath1nyi ) entanglement entropy for this excited state
is given by @xmath327 where @xmath328 is the ( r@xmath1nyi ) entanglement entropy of @xmath329 .
in this section we would like to explain the new results which we obtain in present paper .
firstly we find the explanation which shows that the ( r@xmath1nyi ) entanglement entropies of@xmath3 are given by those of binomial distribution by the replica method .
and we will also obtain the sum rule which the ( r@xmath1nyi ) entanglement entropies for the state defined by acting various operators on the ground state .
we assume that these operators are located separately .
moreover we argue that the ( r@xmath1nyi ) entanglement entropies of the specific operators @xmath330 are given by the ( r@xmath1nyi ) entanglement entropies of binomial distribution .
we will also argue that @xmath34 for the state defined by acting various operators on the ground state obey the same sum rule . these results which we obtain by replica trick will agree with the results which we obtained in terms of entangled pair .
finally we will comment on the effects of a conical singularity .
on @xmath59 there is a conical singularity on the boundary of the subsystem a. at first , it seems that this can affect the r@xmath1nyi entanglement entropy for locally excited states
. however it can not affect the entropy for those states as we will explain later . in the free scalar field theories
, we have to calculate propagators to compute the correlation functions of local operators . by taking the @xmath131 limit ,
the calculations of @xmath34 are simplified due to distinctive features of propagators .
we assume the operators located at @xmath331 and @xmath332 on @xmath68 as in fig.1 .
sheeted geometry .
there are @xmath8 blue wavy lines and red wavy lines respectively .
the dots on @xmath333 are the points where operators are located , width=302 ] sheeted geometry .
there are @xmath8 blue wavy lines and red wavy lines respectively .
the dots on @xmath333 are the points where operators are located , width=302 ] only specific propagators on @xmath59 can dominantly contribute to @xmath56 point correlation function of local operators .
we define these dominant propagators on @xmath59 by @xmath334 . as in fig.13 ,
the propagators which are aligned on a circle are dominant propagators . in this figure
, it shows the configuration of propagators on @xmath333 .
this circle is made of @xmath56 propagators . and
these @xmath56 two point functions necessarily include the propagators between two points on same sheet .
these propagators respectively correspond to the @xmath8 red wavy lines in fig.13 . and the other dominant propagators are those between @xmath335 and @xmath336 @xmath337 they respectively correspond to the @xmath8 blue wavy lines in fig.13
. the other propagators on @xmath59 can not contribute to @xmath56 point functions .
here we assume that operator is located at @xmath338 .
in other words , we explain the features of propagators in the region where @xmath34 increase nontrivially .
even if we take the @xmath131 and take @xmath339 , then these dominant propagators can contribute to @xmath34 .
the ratios of @xmath334 to @xmath340 does not depend on the number of sheets and dimension of spacetime .
this ratio is given by @xmath341 by using this property of dominant propagators , we can acquire the final value @xmath300 for @xmath3 in any dimensions .
there are two contributions to the @xmath56 point correlation function of @xmath3 on @xmath59 .
one contribution corresponds to the product of propagators between two points on same sheet .
this corresponds to the red wavy lines in the right figure of fig.13 .
another one corresponds to the product of propagators between @xmath335 and @xmath336 .
this corresponds to the blue wavy lines in the right figure of fig.13 .
then this @xmath56 point function of @xmath3 on @xmath59 is given by @xmath342 the denominator @xmath343 is given by @xmath344 they are proportional to @xmath345 respectively .
then @xmath340 is cancelled out when we compute @xmath34 .
therefore @xmath300 does not depend on the detail of @xmath6 point function of @xmath69 .
in any dimensions , @xmath246 is given by @xmath346}.\ ] ] in previous paper @xcite , we obtained the ( r@xmath1nyi ) entanglement entropies of @xmath3 in terms of entangled pair .
however we did not obtain them for any k and n by replica method .
in present paper , we obtain it for any k and
n by using the distinctive property of propagators .
moreover the result in ( [ rgfo ] ) agrees with that which we obtained in terms of the entangled pairs .
the relation between @xmath334 and @xmath340 does not depend on the details of the propagators .
therefore , it is expected that that relation can be generalized to that for propagators of general fields in free massless scalar field theory .
we would like to comment on the time evolution for @xmath34 for the states defined by acting @xmath3 on the ground state .
here we assume that operator is located at @xmath347 . in @xmath8 dimension ,
we are able to find the time dependence of the dominant propagators
. its time dependence does not depend on the number of sheets in the @xmath131 limit . in the @xmath131 limit
, the dominant propagators between two points on the same sheet is given by @xmath348 in this limit , the dominant propagators between @xmath349 and @xmath350 is given by @xmath351 we point out a new relation called sum rule by the replica method .
we define the particular operators @xmath353 by @xmath247 . here
@xmath354 means the position of composite operators and @xmath4 means the powers of @xmath69 .
operators @xmath355 act on the ground state .
these operators are located separately .
we can find out the final value of @xmath34 in the @xmath131 limit by the replica method . in present case
there are also the correlations between operators which are located separately . at first
, it seems that they can contribute to the ( r@xmath1nyi ) entanglement entropies for the locally excited states .
however , only dominant propagators can contribute to @xmath34 if we take the @xmath131 limit . in this limit ,
the leading term of dominant propagators are given by @xmath356 where @xmath357 .
on the other hand , the leading term of the propagators between operators which are located separately is much smaller than that of dominant propagators .
therefore these correlation functions can not contribute to @xmath34 .
we are able to explain why only specific propagator can contribute to @xmath34 as follows . here
we define the replica fields of @xmath358 and its copy @xmath185 on each sheet @xmath354 by @xmath359 and @xmath360 respectively @xmath361 .
on @xmath68 , @xmath24 is a regularization parameter for the correlation function of @xmath190 and its copy @xmath362 as in fig.14 .
the distance between @xmath363 and @xmath364 greatly depends on the parameter @xmath24 . on the other hand ,
the distance between @xmath363 and @xmath365 does not depend on @xmath24 compared to the distance between @xmath363 and @xmath364 . here
@xmath366 is put on the different point from the location of @xmath190 on @xmath68 .
the distance between @xmath363 and @xmath364 shrinks compared to the distance between @xmath363 and @xmath365 .
therefore only dominant propagators can be greatly enhanced compared to the other propagators when we take @xmath131 limit . in any dimensions
, @xmath34 is given by the sum of the ( r@xmath1nyi ) entanglement entropy @xmath367 of each operators , @xmath368 and @xmath369 also obey the similar relation , @xmath370 [ reg ] .
the distance @xmath229 between @xmath371 and @xmath366 is much larger than the distance @xmath372 between @xmath371 and @xmath373.,title="fig:",width=302 ] in previous subsections , we obtained the ( r@xmath1nyi ) entanglement entropies of @xmath3 and the sum rule which they obey in the replica method .
only dominant propagators can contribute to them and the computation of @xmath34 is very simplified in the @xmath131 limit . in that argument about propagators , we do not use the property of the propagators of @xmath69 .
therefore we argue that the propagators of derivatives of scalar such as @xmath374 which are aligned on the circle dominantly contribute to the @xmath56 point correlation functions of them .
we define the propagators of these fields on @xmath59 by @xmath375 .
these propagators are given by @xmath376 if the excited state is defined by acting operators which composed of only single - species operator such as @xmath377 on the ground state , then the final value of that entropy for these excited states is given by ( [ rgfo ] ) in any dimensions .
for example , if the local operator is @xmath378 , then the final value of @xmath11-th r@xmath1nyi entanglement entropy is given by @xmath379}.\ ] ] we calculate @xmath380 for the state defined by acting local operators @xmath378 on the ground state directly and we check they agrees with the result in ( [ rphi ] ) .
furthermore the results which we obtained by the replica method are consistent with that in ( [ rphi ] ) .
in turn we prepare the excited state defined by acting various operators on the ground state such as @xmath381 where @xmath319 are general local operators and they are located separately . the index @xmath354 denotes the location of operators . in this case
, it seems that the correlation between @xmath382 and @xmath383 can contribute to @xmath34 .
however the magnitudes of these propagators are also much smaller than those of dominant propagators when we take the @xmath131 limit .
therefore we argue that in any dimensions , ( r@xmath1nyi ) entanglement entropy for the state in ( [ mst ] ) should obey @xmath384 where @xmath385 is the entropy for the state excited by @xmath386 . and
the final value of these entropy also obey the similar rule to that in ( [ msrule ] ) .
we constructed the @xmath11-sheeted geometry by gluing region a on a sheet to the subsystem a on next sheet as in fig.2 .
there is a conical singularity on the boundary @xmath250 of the region a on the n - sheeted geometry . in our case
, it is located at @xmath387 .
this conical singularity deforms the action of the free massless scalar field in more than @xmath6 dimension .
it adds the conformal mass @xmath388 to that action , @xmath389 where @xmath390 is a some constant .
this conformal mass can make an effect on ( r@xmath1nyi ) entanglement entropies if the state is excited .
therefore we have to take its effect into account .
we treat this conformal mass as operator .
if we take the effect of the conical singularity in account , the @xmath56 point correlation function of operators on @xmath59 is given by @xmath391 the second term of the last line in ( [ coni ] ) comes from the conical singularity . in this term
, some of @xmath56 operators has to contract with the operators at @xmath392 .
the propagators between the operator at origin and operators at other point weakly depend on @xmath24 compared to dominant propagators .
therefore when we take @xmath131 limit , the second term in ( [ coni ] ) can not contribute to leading term of @xmath56 point correlation function of @xmath16 .
we can ignore the effect of the conical singularity on the boundary of the region a. by using the simplest example , we explain why the conical singularity can not contribute to the @xmath56 point function of @xmath16 .
we compute the second r@xmath1nyi entanglement entropy for @xmath393 .
the @xmath8 point correlation function of @xmath69 on @xmath152 is given by @xmath394 the correlation function between the origin and the locations of operators are given by @xmath395 the contribution from this correlation function is at most given by @xmath396 after performing the analytic continuation , the contribution from this propagator is at most @xmath397 in @xmath131 limit and @xmath398 is regularized .
therefore the leading term of @xmath399 is @xmath400 and the second term in ( [ n2con ] ) is at most @xmath401 in this limit
. then it can not contribute to the 4 point correlation function of @xmath69 .
therefore we can ignore the effect of the conical singularity .
in the present paper , we studied the ( r@xmath1nyi ) entanglement entropy for a new class of excited states .
these excited states are defined by acting various local operators on the ground state .
therefore we can investigate the evolution of ( r@xmath1nyi ) entanglement entropies with time .
furthermore we defined the excess of ( r@xmath1nyi ) entanglement entropies for those locally excited states .
we chose a half @xmath70 of the total space as the subsystem a in present case .
those entropies are related to the correlation functions of the operators . in general , it is difficult to compute the ( r@xmath1nyi ) entanglement entropy in interacting filed theories . for this excess of ( r@xmath1nyi ) entanglement entropies ,
all that we have to do is to calculate the correlation functions of the operators if we compute them perturbatively .
it is expected that this will be perturbatively computable in interacting filed theory . by the replica method we computed ( r@xmath1nyi ) entanglement entropies for excited states defined by acting particular operators on the ground state in free massless scalar field theory in even dimension .
we performed an analytic continuation to real time and investigated their time evolution .
the ( r@xmath1nyi ) entanglement entropies for locally excited states approach some finite constants at late time .
we found their time evolution can be interpreted in terms of relativistic propagation of the entangled pair .
we gave an explanation of the final value of the ( r@xmath1nyi ) entanglement entropies for those locally excited states in terms of entangled pair interpretation .
we defined the ( r@xmath1nyi ) entanglement entropies of operators by the final value of the ( r@xmath1nyi ) entanglement entropies for those locally excited states .
they characterize operators in the view point of quantum entanglement and they are independent of the conformal dimension of operators .
the ( r@xmath1nyi ) entanglement entropies of @xmath3 are given by ( [ gfo ] ) .
they do not depend on the spacetime dimension .
they do not change if we deform the shape of subsystem a continuously .
they can be defined in any dimensions .
we obtained various new results that we summarize below . in previous paper @xcite , we computed the n - th r@xmath1nyi entanglement entropies for the states which are excited by @xmath3 by replica method when @xmath402 is small . in this paper
, we obtained the explanation which shows that the ( r@xmath1nyi ) entanglement entropies of@xmath3 are given by those of binomial distribution by the replica trick for any @xmath11 and @xmath4 and any dimensions .
we checked that they agree with the results which we obtained in terms of entangled pair .
we argued the ( r@xmath1nyi ) entanglement entropies of particular operators are also given by those of binomial distribution .
these operators are composed of single - species operators such as @xmath403 and we call them s - operators
. we studied the ( r@xmath1nyi ) entanglement entropies of @xmath3 in large @xmath4 limit .
we also obtained the sum rule for the final value of the ( r@xmath1nyi ) entanglement entropies for locally excited states in any dimensions . here
those locally excited states were defined by acting various local operators on the ground state .
we assumed these operators were separated spatially and they were given by @xmath3 .
we checked that the sum rule which we obtained by the replica method agrees with the one derived from the entangled pair .
we argued that @xmath369 for the states defined by acting general operators on the ground state obey the same sum rule in any dimension .
here we assumed that those operators were located separately . &
4 & @xmath404 } $ ] + second r@xmath1nyi entropy & 6 & @xmath405}$ ] + & 8 & @xmath406}$ ] + we would also like to list a few of open problems .
as we mentioned above , we found the evidence that the ( r@xmath1nyi ) entanglement entropies of s - operators will be given by ( [ gfo ] ) .
let us compute the ( r@xmath1nyi ) entanglement entropies of the operators which are composed of multi - species operators such as @xmath407 , @xmath408 and so on .
they are not s - operators . we call them m - operators .
it is expected that the ( r@xmath1nyi ) entanglement entropies of m - operators will not obey the formula given in ( [ gfo ] ) .
for example , we prepare the state defined by acting an m - operator @xmath409 on the ground state , @xmath410 the ( r@xmath1nyi ) entanglement entropy of @xmath409 depends on the spacetime dimension as in table.2 . the time evolution of @xmath102 for this state is plotted in fig.15 .
the second r@xmath1nyi entanglement entropy does not obey the formula in ( [ gfo ] ) .
let us interpret it in terms of entangled pair .
we can decompose operators to the left moving modes and right moving modes respectively , @xmath411 in this case , the relation between @xmath69 , @xmath412 is not clear
. then these operators are probably not independent of each other .
as we explained , the second r@xmath1nyi entanglement entropy of @xmath409 depends on the spacetime dimension and does not obey the formula in ( [ gfo ] ) .
therefore , it is interesting to understand why @xmath102 depends on spacetime dimensions and to interpret it physically .
it is one of the open problems .
[ l10dep ] with @xmath35 in the @xmath131 limit .
the vertical axis corresponds to @xmath102 .
the horizontal axis corresponds to @xmath35 .
here we chose @xmath140 .
the green curve is the time evolution of @xmath102 for @xmath413 in @xmath8 dimension .
the blue curve is the time evolution of @xmath102 for that state in @xmath10 dimension .
the red curve is the time evolution of @xmath102 for that state in @xmath207 dimension .
, title="fig:",width=302 ] other interesting open problems include : * it is interesting to find the formula for the ( r@xmath1nyi ) entanglement entropies of m - operators * it is interesting to investigate the time evolution of ( r@xmath1nyi ) entanglement entropy for the locally excited states in interacting field theories .
* it will be also interesting to investigate the ( r@xmath1nyi ) entanglement entropies of operators in massive theory .
also we would like to refer to @xcite for large n cft and refer to @xcite for rational cfts in two dimension .
mn would like to thank the collaborators t. takayanagi , and t. numasawa for reading the draft of this paper and giving us very useful and remarkable comments .
mn also thanks p. caputa , s. he and n. shiba for reading the draft of this paper and giving us useful and remarkable comments .
mn is supported by jsps fellowship .
h. li and f. d. m. haldane , `` entanglement spectrum as a generalization of entanglement entropy : identification of topological order in non - abelian fractional quantum hall effect states , '' phys .
rev.lett .
* 101 * , 010504 ( 2008 ) [ arxiv:0805.0332 [ cond-mat.mes-hall ] ] .
m. b. hastings , i. gonzalez , a. b. kallin , and r. g. melko , `` measuring renyi entanglement entropy in quantum monte carlo simulations , '' phys.rev . lett . * 104 * , 157201 ( 2010 ) [ arxiv:1001.2335 [ cond-mat.str-el ] ] .
b. swingle , `` entanglement renormalization and holography , '' phys .
d * 86 * , 065007 ( 2012 ) [ arxiv:0905.1317 [ cond-mat.str-el ] ] .
b. swingle , `` constructing holographic spacetimes using entanglement renormalization , '' arxiv:1209.3304 [ hep - th ]
. m. van raamsdonk , `` building up spacetime with quantum entanglement , '' gen .
grav . * 42 * , 2323 ( 2010 ) [ int . j. mod
. phys .
d * 19 * , 2429 ( 2010 ) ] [ arxiv:1005.3035 [ hep - th ] ] ; m. van raamsdonk , `` comments on quantum gravity and entanglement , '' arxiv:0907.2939 [ hep - th ] .
s. ryu and t. takayanagi , `` aspects of holographic entanglement entropy , '' jhep * 0608 * , 045 ( 2006 ) [ hep - th/0605073 ] ; s. ryu and t. takayanagi , `` holographic derivation of entanglement entropy from ads / cft , '' phys . rev . lett . *
96 * , 181602 ( 2006 ) [ hep - th/0603001 ] .
m. nozaki , s. ryu and t. takayanagi , `` holographic geometry of entanglement renormalization in quantum field theories , '' jhep * 1210 * , 193 ( 2012 ) [ arxiv:1208.3469 [ hep - th ] ] ; a. mollabashi , m. nozaki , s. ryu and t. takayanagi , `` holographic geometry of cmera for quantum quenches and finite temperature , '' jhep * 1403 * , 098 ( 2014 ) [ arxiv:1311.6095 [ hep - th ] ] .
m. nozaki , t. numasawa , a. prudenziati and t. takayanagi , `` dynamics of entanglement entropy from einstein equation , '' phys .
d * 88 * , 026012 ( 2013 ) [ arxiv:1304.7100 [ hep - th ] ] ; j. bhattacharya and t. takayanagi , `` entropic counterpart of perturbative einstein equation , '' jhep * 1310 * , 219 ( 2013 ) [ arxiv:1308.3792 [ hep - th ] ] .
t. faulkner , m. guica , t. hartman , r. c. myers and m. van raamsdonk , `` gravitation from entanglement in holographic cfts , '' jhep * 1403 * , 051 ( 2014 ) [ arxiv:1312.7856 [ hep - th ] ] ; n. lashkari , m. b. mcdermott and m. van raamsdonk , `` gravitational dynamics from entanglement `` thermodynamics '' , '' jhep * 1404 * , 195 ( 2014 ) [ arxiv:1308.3716 [ hep - th ] ] .
see , for example : j. bhattacharya , m. nozaki , t. takayanagi and t. ugajin , `` thermodynamical property of entanglement entropy for excited states , '' phys .
lett . * 110 * , no . 9 , 091602 ( 2013 ) [ arxiv:1212.1164 ] ; d. d. blanco , h. casini , l. -y
. hung and r. c. myers , `` relative entropy and holography , '' jhep * 1308 * , 060 ( 2013 ) [ arxiv:1305.3182 [ hep - th ] ] ; d. allahbakhshi , m. alishahiha and a. naseh , `` entanglement thermodynamics , '' jhep * 1308 * , 102 ( 2013 ) [ arxiv:1305.2728 [ hep - th ] ] .
m. nozaki , t. numasawa and t. takayanagi , `` holographic local quenches and entanglement density , '' jhep * 1305 * , 080 ( 2013 ) [ arxiv:1302.5703 [ hep - th ] ] .
w. -z .
guo , s. he and j. tao , `` note on entanglement temperature for low thermal excited states in higher derivative gravity , '' jhep * 1308 * , 050 ( 2013 ) [ arxiv:1305.2682 [ hep - th ] ] ; a. bhattacharyya , a. kaviraj and a. sinha , `` entanglement entropy in higher derivative holography , '' jhep * 1308 * , 012 ( 2013 ) [ arxiv:1305.6694 [ hep - th ] ] ; s. he , d. li and j. -b .
wu , `` entanglement temperature in non - conformal cases , '' jhep * 1310 * , 142 ( 2013 ) [ arxiv:1308.0819 [ hep - th ] ] ; p. caputa , v. jejjala and h. soltanpanahi , `` entanglement entropy of extremal btz , '' phys .
d * 89 * , 046006 ( 2014 ) [ arxiv:1309.7852 [ hep - th ] ] ; d. -w .
pang , `` entanglement thermodynamics for nonconformal d - branes , '' phys .
d * 88 * ( 2013 ) 12 , 126001 [ arxiv:1310.3676 [ hep - th ] ] ; a. bhattacharyya , m. sharma and a. sinha , `` on generalized gravitational entropy , squashed cones and holography , '' jhep * 1401 * , 021 ( 2014 ) [ arxiv:1308.5748 [ hep - th ] ] ; a. bhattacharyya and m. sharma , `` on entanglement entropy functionals in higher derivative gravity theories , '' arxiv:1405.3511 [ hep - th ] .
see , for example : p. calabrese and j. l. cardy , `` evolution of entanglement entropy in one - dimensional systems , '' j. stat .
* 04 * ( 2005 ) p04010 , cond - mat/0503393 ; p. calabrese and j. l. cardy , `` time - dependence of correlation functions following a quantum quench , '' phys.rev.lett.*96*(2006)136801 , cond - mat/0601225 ; p. calabrese and j. l. cardy , `` quantum quenches in extended systems , '' j. stat .
mech . * 06 * ( 2007 ) p06008 , arxiv:0704.1880 ; s. sotiriadis and j. cardy , `` quantum quench in interacting field theory : a self - consistent approximation , '' phys . rev .
b * 81 * ( 2010 ) 134305 , [ arxiv:1002.0167 [ quant - ph ] ] .
s. bhattacharyya and s. minwalla , `` weak field black hole formation in asymptotically ads spacetimes , '' jhep * 0909 * ( 2009 ) 034 [ arxiv:0904.0464 [ hep - th ] ] ; s. r. das , t. nishioka and t. takayanagi , `` probe branes , time - dependent couplings and thermalization in ads / cft , '' jhep * 1007 * ( 2010 ) 071 [ arxiv:1005.3348 [ hep - th ] ] ; h. ebrahim and m. headrick , `` instantaneous thermalization in holographic plasmas , '' arxiv:1010.5443 [ hep - th ] ; d. garfinkle and l. a. pando zayas , `` rapid thermalization in field theory from gravitational collapse , '' phys . rev .
d * 84 * ( 2011 ) 066006 [ arxiv:1106.2339 [ hep - th ] ] ; p. basu and s. r. das , `` quantum quench across a holographic critical point , '' jhep * 1201 * ( 2012 ) 103 [ arxiv:1109.3909 [ hep - th ] ] ; s. r. das , `` holographic quantum quench , '' j. phys
. conf .
* 343 * ( 2012 ) 012027 [ arxiv:1111.7275 [ hep - th ] ] ; e. caceres and a. kundu , `` holographic thermalization with chemical potential , '' jhep * 1209 * ( 2012 ) 055 [ arxiv:1205.2354 [ hep - th ] ] ; a. buchel , l. lehner and r. c. myers , `` thermal quenches in n=2 * plasmas , '' jhep * 1208 * ( 2012 ) 049 [ arxiv:1206.6785 [ hep - th ] ] ; m. j. bhaseen , j. p. gauntlett , b. d. simons , j. sonner and t. wiseman , `` holographic superfluids and the dynamics of symmetry breaking , '' arxiv:1207.4194 [ hep - th ] .
p. basu , d. das , s. r. das and t. nishioka , `` quantum quench across a zero temperature holographic superfluid transition , '' arxiv:1211.7076 [ hep - th ] .
x. gao , a. m. garcia - garcia , h. b. zeng and h. -q .
zhang , `` lack of thermalization in holographic superconductivity , '' arxiv:1212.1049 [ hep - th ] . w. h. baron , d. galante and m. schvellinger , `` dynamics of holographic thermalization , '' arxiv:1212.5234 [ hep - th ] ; a. buchel , l. lehner , r. c. myers and a. van niekerk , `` quantum quenches of holographic plasmas , '' arxiv:1302.2924 [ hep - th ] ; t. takayanagi and t. ugajin , `` measuring black hole formations by entanglement entropy via coarse - graining , '' jhep * 1011 * ( 2010 ) 054 [ arxiv:1008.3439 [ hep - th ] ] .
t. hartman and j. maldacena , `` time evolution of entanglement entropy from black hole interiors , '' jhep * 1305 * , 014 ( 2013 ) [ arxiv:1303.1080 [ hep - th ] ] .
f. c. alcaraz , m. i. berganza and g. sierra , `` entanglement of low - energy excitations in conformal field theory , '' phys .
lett . *
106 * , 201601 ( 2011 ) [ arxiv:1101.2881 [ cond-mat.stat-mech ] ] .
a. f. astaneh and a. e. mosaffa , `` holographic entanglement entropy for excited states in two dimensional cft , '' jhep * 1303 * , 135 ( 2013 ) [ arxiv:1301.1495 [ hep - th ] ] .
s. ryu and t. takayanagi , `` aspects of holographic entanglement entropy , '' jhep * 0608 * , 045 ( 2006 ) [ hep - th/0605073 ] .
p. calabrese and j. l. cardy , `` entanglement and correlation functions following a local quench : a conformal field theory approach , '' j.phys.a42:504005,2009 , [ arxiv:0905.4013 [ cond-mat.stat-mech ] ] .
j. l. cardy , o. a. castro - alvaredo and b. doyon , `` form factors of branch - point twist fields in quantum integrable models and entanglement entropy , '' j. statist .
phys . *
130 * , 129 ( 2008 ) [ arxiv:0706.3384 [ hep - th ] ] . p. calabrese and j. l. cardy , `` entanglement entropy and quantum field theory , '' j. stat .
mech . * 0406 * , p06002 ( 2004 ) [ hep - th/0405152 ] .
metlitski , max a. fuertes , carlos a. and sachdev , subir `` entanglement entropy in the @xmath414 model '' phys.rev.b.80,*115122*(2009 ) [ hep - th/0904.4477 ] s. deser and r. jackiw , `` classical and quantum scattering on a cone , '' commun .
phys . * 118 * , 495 ( 1988 ) .
m. e. x. guimaraes and b. linet , `` scalar green s functions in an euclidean space with a conical - type line singularity , '' commun .
phys . * 165 * , 297 ( 1994 ) .
j. s. dowker , `` quantum field theory on a cone , '' j. phys .
a * 10 * , 115 ( 1977 ) .
s. he , t. numasawa , t. takayanagi and k. watanabe , `` quantum dimension as entanglement entropy in 2d cfts , '' arxiv:1403.0702 [ hep - th ] .
a. lewkowycz and j. maldacena , `` exact results for the entanglement entropy and the energy radiated by a quark , '' arxiv:1312.5682 [ hep - th ] .
p. cupta , m. nozaki , t. takayanagi `` entanglement of local operators in large n cfts , '' to appear . | this is an expanded version of the short report arxiv:1401.0539 , where we studied the ( renyi ) entanglement entropies for the excited state defined by acting a given local operator on the ground state .
we introduced the ( renyi ) entanglement entropies of given local operators which measure the degrees of freedom of local operators and characterize them in conformal field theories from the viewpoint of quantum entanglement .
in present paper , we explain how to compute them in free massless scalar field theories and we also investigate their time evolution .
the results are interpreted in terms of relativistic propagation of an entangled pair .
the main new results which we acquire in the present paper are as follows .
firstly , we provide an explanation which shows that the ( renyi ) entanglement entropies of a specific operator are given by ( renyi ) entanglement entropies of binomial distribution by the replica method .
that operator is constructed of only scalar field .
secondly , we found the sum rule which ( renyi ) entanglement entropies of those local operators obey .
those local operators are located separately
. moreover we argue that ( renyi ) entanglement entropies of specific operators in conformal field theories are given by ( renyi ) entanglement entropies of binomial distribution .
these specific operators are constructed of single - species operator .
we also argue that general operators obey the sum rule which we mentioned above .
yitp-14 - 41 * notes on quantum entanglement of local operators * + masahiro nozaki @xmath0 _ @xmath0yukawa institute for theoretical physics , kyoto university , kyoto 606 - 8502 , japan + _ 2em |
a crucial question in elementary particle physics is whether nature makes use of the spontaneous symmetry breaking for generating the masses or not . in the minimal standard model ( sm )
this mechanism requires the existence of a scalar particle , the higgs boson .
extensions of the sm , e.g. models with more than one higgs doublet or supersymmetric versions of the sm , predict also pseudo - scalar higgs bosons , @xmath5 .
however , up to now there is no experimental evidence for such a particle and the direct search at lep via the process @xmath6 rules out the mass range @xmath7 gev with a 95% confidence level ( cl ) @xcite .
assuming the validity of the sm the precision data are sensitive to the higgs boson and a recent global fit yields @xmath8 gev together with a 95% cl upper bound of 550 gev @xcite ( in this context see also @xcite ) .
theoretical arguments based on unitarity or the validity of perturbation theory request an upper limit on the higgs mass of about 1 tev . in this paper
we consider the decay of a higgs boson , scalar or pseudo - scalar , with mass above 400 gev , into a top - antitop quark pair . since the velocity of one of the quarks for this mass range is @xmath9 , threshold effects can safely be neglected and the decay rate can be treated in a purely perturbative way . the full mass dependence for this process is only known to @xmath10 @xcite ,
while to order @xmath11 only mass terms up to @xmath12 are available @xcite , where @xmath13 is the top quark mass , and they show up to be quite sizeable @xcite .
@xmath14 corrections have recently been evaluated in the massless case @xcite . in the following , making use of a technique recently developed for the automatic computation of mass corrections to the vector current correlator @xcite
, we will demonstrate that the inclusion of mass terms up to order @xmath15 leads to reliable results at @xmath1 in the considered mass range .
we note that the results presented in this paper for the decay into top quarks may be generalized to any fermion species .
however , even for bottom quarks already the quadratic mass corrections are very small . to fix the notation we define : @xmath16 where @xmath17 ( @xmath18 ) represents the scalar ( pseudo - scalar ) current correlator in momentum space .
the currents are given by @xmath19 it should be noted that as far as renormalization is concerned it is convenient to consider the combination @xmath20 in the on - shell scheme and @xmath21 in the @xmath22 scheme , where @xmath23 ( @xmath24 ) is the generic pole ( @xmath22 ) mass , in order to avoid additional renormalization constants .
the physical observable @xmath25 is related to @xmath26 via the relation @xmath27 the current correlator can be written as @xmath28 and similarly for @xmath25 .
the normalization in eq .
( [ eqrtopisp ] ) guarantees that @xmath29 for @xmath30 .
the colour factors ( @xmath31 and @xmath32 ) correspond to the casimir operators of the fundamental and adjoint representations , respectively . for the numerical evaluation we set @xmath33 .
the trace normalization of the fundamental representation is @xmath34 .
the number of light ( massless ) quark flavours is denoted by @xmath35 . in eq .
( [ eqpi2 ] ) @xmath36 is the abelian contribution which also exists in qed , and @xmath37 results from the non - abelian structure specific for qcd . the contribution of diagrams containing a second massless or massive quark loop is denoted by @xmath38 and @xmath39 , respectively .
@xmath40 represents the terms arising from the double - triangle diagram and is called the singlet contribution . the case for the double - bubble diagram where the inner quark mass is much heavier than the outer one
is not listed as it does not contribute in cases of physical interest .
the outline of the paper is as follows : in section [ seclmp ] the method is briefly described and the results for the correlator functions are given . in section [ sechtt ] mass effects for the decay of a scalar and pseudo - scalar higgs particle into top quarks
are presented and the results are discussed .
in this section we keep the discussion general and consider the scalar and pseudo - scalar current correlators with a generic quark mass @xmath23 . in section [ sechtt ] the specification to the case @xmath41 is performed and numerical values are given .
let us in a first step outline the main ideas of the large momentum expansion and its realization by a computer .
the large momentum procedure @xcite requires the identification of certain subgraphs associated with the feynman diagram to be calculated .
these subgraphs then have to be expanded in their small dimensional quantities ( i.e. all except the large momenta ) and the resulting terms , being products of tadpoles and massless integrals , have to be calculated . the number of these terms increases rapidly with the number of loops . in our case of a massive current correlator ,
the one- and two - loop diagrams generate 17 terms altogether , so that they still can be treated by hand . in the three - loop case
, however , there are 19 topologies , contributing 266 terms , and their generation by hand is not feasible .
therefore we completely automated the large momentum procedure for massive two - point functions and directly fed the generated terms to the form @xcite packages matad and mincer @xcite . after calculating the single terms the results of all diagrams
were added giving a finite expression after renormalization . we are now prepared to present the result .
let us first consider the scalar correlator .
the results for the different contributions read in the @xmath22 scheme ( @xmath42 : @xmath43 + \left({\bar{m}^2\over q^2}\right)^{2 } \ , \bigg [ - 18 - 12\,{l_{qm}}\bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ { 4\over 3 } - 8\,{l_{qm}}\bigg ] + \left({\bar{m}^2\over q^2}\right)^{4 } \ , \bigg [ 7 - 12\,{l_{qm}}\bigg ] \bigg\ } + \ldots\,\ , , \label{eqpi0s } \\[.4 cm ] \bar{\pi}^{(1),s } & = & { 3\over 16 \pi^2 } \bigg\ { { 131\over 8 } - 6\,\zeta_3 - { 17\over 2}\,{l_{q\mu}}+ { 3\over 2}\,{l_{q\mu}}^2 \nonumber\\&&\mbox { } + { \bar{m}^2\over q^2}\ , \bigg [ - 94 + 36\,\zeta_3 + 60\,{l_{q\mu}}- 18\,{l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{2 } \ , \bigg [ - 44 - 48\,\zeta_3 - 69\,{l_{qm}}- 18\,{l_{qm}}^2 + \left(63 + 54\,{l_{qm}}\right ) { l_{q\mu}}\bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ - { 461\over 9 } - { 956\over 9}\,{l_{qm}}- { 98\over 3}\,{l_{qm}}^2 + \left(- 20 + 48\,{l_{qm}}\right ) { l_{q\mu}}\bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{4 } \ , \bigg [ { 8263\over 144 } - { 881\over 6}\,{l_{qm}}- { 135\over 2}\,{l_{qm}}^2 + \left(- { 141\over 2 } + 90\,{l_{qm}}\right ) { l_{q\mu}}\bigg ] \bigg\ } + \ldots\,\ , , \label{eqpi1s } \\[.4 cm ] \bar{\pi}^{(2),s}_{\it a } & = & { 3\over 16 \pi^2 } \bigg\ { { 1613\over 64 } - 24\,\zeta_3 + { 9\over
4}\,\zeta_4 + 15\,\zeta_5 + \left(- { 691\over 32 } + { 9\over 2}\,\zeta_3 \right ) { l_{q\mu}}+ { 105\over 16}\,{l_{q\mu}}^2 - { 3\over 4}\,{l_{q\mu}}^3 \nonumber\\&&\mbox { } + { \bar{m}^2\over q^2}\ , \bigg [ - { 3471\over 16 } + 220\,\zeta_3 - 18\,\zeta_4 - 130\,\zeta_5 \nonumber\\&&\mbox{\hspace{.2 cm } } + \left({1911\over 8 } - 72\,\zeta_3 \right ) { l_{q\mu}}- { 369\over 4}\,{l_{q\mu}}^2 + 18\,{l_{q\mu}}^3 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{2 } \ , \bigg [ - { 4517\over 32 } - 340\,\zeta_3 + 72\,\zeta_4 + 115\,\zeta_5 - 12\,b_4 \nonumber\\&&\mbox{\hspace{.2 cm } } + \left(- { 2871\over 16 } + 27\,\zeta_3 \right ) { l_{qm}}- { 315\over 4}\,{l_{qm}}^2 - 18\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left({819\over 8 } + 216\,\zeta_3 + { 1053\over 4}\,{l_{qm}}+ 81\,{l_{qm}}^2 \right ) { l_{q\mu}}+ \left(- { 405\over 4 } - { 243\over 2}\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ - { 48895\over 1944 } - 427\,\zeta_3 + 48\,\zeta_4 - 80\,\zeta_5 - 8\,b_4 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left(- { 8938\over 27 } - 116\,\zeta_3 \right ) { l_{qm}}- { 1375\over 6}\,{l_{qm}}^2 - { 1639\over 27}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left({291\over 2 } + { 1636\over 3}\,{l_{qm}}+ 196\,{l_{qm}}^2 \right ) { l_{q\mu}}+ \left(96 - 144\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{4 } \ , \bigg [ - { 6460859\over 31104 } - 60\,\zeta_3 + 72\,\zeta_4 - 40\,\zeta_5 - 12\,b_4 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left(- { 1242845\over 1296 } - 173\,\zeta_3 \right ) { l_{qm}}- { 546029\over 864}\,{l_{qm}}^2 - { 32735\over 216}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left(- { 63305\over 96 } + 910\,{l_{qm}}+ { 2025\over 4}\,{l_{qm}}^2 \right ) { l_{q\mu}}+ \left({2655\over 8 } - { 675\over 2}\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \bigg\ } + \ldots\,\ , , \label{eqpi2as } \\[.4 cm ] \bar{\pi}^{(2),s}_{\it na } & = & { 3\over 16 \pi^2 } \bigg\ { { 14419\over 288 } - { 75\over 4}\,\zeta_3
- { 9\over 8}\,\zeta_4 - { 5\over 2}\,\zeta_5 + \left(- { 893\over 32 } + { 31\over 4}\,\zeta_3 \right ) { l_{q\mu}}+ { 71\over 12}\,{l_{q\mu}}^2 - { 11\over 24}\,{l_{q\mu}}^3 \nonumber\\&&\mbox { } + { \bar{m}^2\over q^2}\ , \bigg [ - { 12617\over 48 } + 90\,\zeta_3 + 9\,\zeta_4 + 15\,\zeta_5 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left({4601\over 24 } - 51\,\zeta_3 \right ) { l_{q\mu}}- { 207\over 4}\,{l_{q\mu}}^2 + { 11\over 2}\,{l_{q\mu}}^3 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{2 } \ , \bigg [ - { 50347\over 288 } + { 467\over 12}\,\zeta_3 - 36\,\zeta_4 - { 95\over 2}\,\zeta_5 + 6\,b_4 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left(- { 9503\over 48 } + { 27\over 2}\,\zeta_3 \right ) { l_{qm}}- 38\,{l_{qm}}^2 - { 11\over 2}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left({3005\over 24 } + 44\,\zeta_3 + 136\,{l_{qm}}+ { 33\over 2}\,{l_{qm}}^2 \right ) { l_{q\mu}}+ \left(- { 231\over 8 } - { 99\over 4}\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ - { 61601\over 972 } + { 7\over 2}\,\zeta_3 - 24\,\zeta_4 + 20\,\zeta_5 + 4\,b_4 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left(- { 208093\over 648 } + 8\,\zeta_3 \right ) { l_{qm}}- { 563\over 6}\,{l_{qm}}^2 - { 521\over 54}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left({2161\over 108 } + { 4375\over 27}\,{l_{qm}}+ { 539\over 18}\,{l_{qm}}^2 \right ) { l_{q\mu}}+
\left({55\over 6 } - 22\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{4 } \ , \bigg [ { 396607\over 1296 } - { 1393\over 24}\,\zeta_3 - 36\,\zeta_4
+ 10\,\zeta_5 + 6\,b_4 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left(- { 368443\over 864 } + 63\,\zeta_3 \right ) { l_{qm}}- { 57929\over 288}\,{l_{qm}}^2 - { 803\over 24}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left(- { 255017\over 1728 } + { 18421\over 72}\,{l_{qm}}+ { 495\over 8}\,{l_{qm}}^2 \right ) { l_{q\mu}}+
\left({517\over 16 } - { 165\over 4}\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \bigg\ } + \ldots\,\ , , \nonumber\\ \label{eqpi2nas } \\[.4 cm ] \bar{\pi}^{(2),s}_{\it l } & = & { 3\over 16 \pi^2 } \bigg\ { - { 511\over 36 } + 4\,\zeta_3 + \left({65\over 8 } - 2\,\zeta_3 \right ) { l_{q\mu}}- { 11\over 6}\,{l_{q\mu}}^2 + { 1\over 6}\,{l_{q\mu}}^3 \nonumber\\&&\mbox { } + { \bar{m}^2\over q^2}\ , \bigg [ { 817\over 12 } - 12\,\zeta_3 + \left(- { 313\over 6 } + 12\,\zeta_3 \right ) { l_{q\mu}}+ 15\,{l_{q\mu}}^2 - 2\,{l_{q\mu}}^3 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{2 } \ , \bigg [ { 3311\over 72 } - { 4\over 3}\,\zeta_3 + { 595\over 12}\,{l_{qm}}+ 10\,{l_{qm}}^2 + 2\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left(- { 193\over 6 } - 16\,\zeta_3 - 38\,{l_{qm}}- 6\,{l_{qm}}^2 \right ) { l_{q\mu}}+
\left({21\over 2 } + 9\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ - { 8347\over 486 } + { 80\over 3}\,\zeta_3 + { 4534\over 81}\,{l_{qm}}+ { 149\over 9}\,{l_{qm}}^2 + { 62\over 27}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left(- { 311\over 27 } - { 1316\over 27}\,{l_{qm}}- { 98\over 9}\,{l_{qm}}^2 \right ) { l_{q\mu}}+ \left(- { 10\over 3 } + 8\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{4 } \ , \bigg [ - { 54461\over 1296 } + 20\,\zeta_3 + { 2821\over 27}\,{l_{qm}}+ { 287\over 8}\,{l_{qm}}^2 + { 25\over 6}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left({16723\over 432 } - { 1331\over 18}\,{l_{qm}}- { 45\over 2}\,{l_{qm}}^2 \right ) { l_{q\mu}}+ \left(- { 47\over 4 } + 15\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \bigg\ } + \ldots\,\ , , \label{eqpi2ls } \\[.4 cm ] \bar{\pi}^{(2),s}_{\it
f } & = & { 3\over 16 \pi^2 } \bigg\ { - { 511\over 36 } + 4\,\zeta_3 + \left({65\over 8 } - 2\,\zeta_3 \right ) { l_{q\mu}}- { 11\over 6}\,{l_{q\mu}}^2 + { 1\over 6}\,{l_{q\mu}}^3 \nonumber\\&&\mbox { } + { \bar{m}^2\over q^2}\ , \bigg [ { 881\over 12 } - 12\,\zeta_3 + \left(- { 385\over 6 } + 12\,\zeta_3 \right ) { l_{q\mu}}+ 15\,{l_{q\mu}}^2 - 2\,{l_{q\mu}}^3 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{2 } \
, \bigg [ { 9863\over 72 } - { 94\over 3}\,\zeta_3 + { 1135\over 12}\,{l_{qm}}+ { 23\over 2}\,{l_{qm}}^2 + 2\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left(- { 193\over 6 } - 16\,\zeta_3 - 38\,{l_{qm}}-
6\,{l_{qm}}^2 \right ) { l_{q\mu}}+ \left({21\over 2 } + 9\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ { 10757\over 486 } - { 628\over 9}\,\zeta_3 + { 4316\over
27}\,{l_{qm}}+ { 1195\over 27}\,{l_{qm}}^2 + { 124\over 27}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left(- { 311\over 27 } - { 1316\over 27}\,{l_{qm}}- { 98\over 9}\,{l_{qm}}^2 \right ) { l_{q\mu}}+ \left(- { 10\over 3 } + 8\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{4 } \ , \bigg [ - { 193831\over 864 } - 187\,\zeta_3 + { 13051\over 72}\,{l_{qm}}+ { 7699\over
72}\,{l_{qm}}^2 + { 47\over 3}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left({16723\over 432 } - { 1331\over 18}\,{l_{qm}}- { 45\over 2}\,{l_{qm}}^2 \right ) { l_{q\mu}}+ \left(- { 47\over 4 } + 15\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \bigg\ } + \ldots\,\ , , \label{eqpi2fs } \\[.4 cm ] \bar{\pi}^{(2),s}_{\it s } & = & { 3\over 16 \pi^2 } \bigg\ { { \bar{m}^2\over q^2}\ , \bigg [ { 236\over 3 } - 40\,\zeta_3 - 20\,\zeta_5 - 24\,{l_{q\mu}}\bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{2 } \ , \bigg [ - 84 + 8\,\zeta_3 + 160\,\zeta_5 + \left(- 36 + 72\,\zeta_3 \right ) { l_{qm}}\bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ { 37\over 8 } - 62\,\zeta_3 + 320\,\zeta_5 + \left(- { 3\over 4 } - 36\,\zeta_3 \right ) { l_{qm}}+ 33\,{l_{qm}}^2 + 12\,{l_{qm}}^3 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{4 } \ , \bigg [ { 178423\over 243 } - { 4472\over 9}\,\zeta_3 + \left({22289\over 81 } + 16\,\zeta_3 \right ) { l_{qm}}\nonumber\\&&\mbox{\hspace{0.4 cm } } - 26\,{l_{qm}}^2 - { 28\over 3}\,{l_{qm}}^3 \label{eqpi2ss } \bigg ] \bigg\ } + \ldots\,\ , , \ ] ] where @xmath44 is riemann s zeta function with values @xmath45 , @xmath46 @xmath47 and @xmath48 . the constant @xmath49 is typical for three - loop tadpole integrals and is given by @xmath50 @xcite .
the corresponding results in the on - shell scheme are obtained by substituting in the combination @xmath51 the @xmath22 mass , @xmath24 , w.r.t .
the on - shell mass @xcite .
for the pseudo - scalar current correlator a minor complication arises in connection with @xmath52 . while the non - singlet contribution allows the use of the anticommuting definition , for the singlet diagram we take the prescription for @xmath52 introduced in @xcite .
thereby we follow the strategy outlined in @xcite and replace @xmath52 in the vertices according to @xmath53},\end{aligned}\ ] ] where @xmath54}$ ] represents the antisymmetrized product of four @xmath55-matrices and may be written as @xmath56 } & = & \frac{1}{4}\left ( \gamma^{\mu}\gamma^{\nu}\gamma^{\rho}\gamma^{\sigma } + \gamma^{\sigma}\gamma^{\rho}\gamma^{\nu}\gamma^{\mu } -\gamma^{\nu}\gamma^{\rho}\gamma^{\sigma}\gamma^{\mu } -\gamma^{\mu}\gamma^{\sigma}\gamma^{\rho}\gamma^{\nu } \right).\end{aligned}\ ] ] because the double - triangle diagram contains no sub - divergencies and consequently has a finite imaginary part it is allowed to contract the new polarization tensor with eight indices with the product of four metric tensors and compute the scalar integrals in the same way as for the non - singlet contributions .
even more , in contrast to the scalar case @xmath57 is finite from the very beginning .
the results for the pseudo - scalar case read : @xmath58 \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ - { 20\over 3 } + 8\,{l_{qm}}\bigg ] + \left({\bar{m}^2\over q^2}\right)^{4 } \ , \bigg [ - { 59\over 3 } + 20\,{l_{qm}}\bigg ] \bigg\ } + \ldots\,\ , , \label{eqpi0p } \\[.4 cm ] \bar{\pi}^{(1),p } & = & { 3\over 16 \pi^2 } \bigg\ { { 131\over 8 } - 6\,\zeta_3 - { 17\over 2}\,{l_{q\mu}}+ { 3\over 2}\,{l_{q\mu}}^2 + { \bar{m}^2\over q^2}\ , \bigg [ - 6 + 12\,\zeta_3 + 4\,{l_{q\mu}}- 6\,{l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{2 } \
, \bigg [ 4 + 27\,{l_{qm}}+ 6\,{l_{qm}}^2 + \left ( 15 - 18\,{l_{qm}}\right ) { l_{q\mu}}\bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ - { 515\over 9 } + { 628\over 9}\,{l_{qm}}+ { 82\over 3}\,{l_{qm}}^2 + \left ( 52 - 48\,{l_{qm}}\right ) { l_{q\mu}}\bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{4 } \ , \bigg [ - { 36985\over 144 } + { 3101\over 18}\,{l_{qm}}+ { 539\over 6}\,{l_{qm}}^2 + \left ( { 355\over 2 } - 150\,{l_{qm}}\right ) { l_{q\mu}}\bigg ] \bigg\ } + \ldots\,\ , , \nonumber\\ \label{eqpi1p } \\[.4 cm ] \bar{\pi}^{(2),p}_{\it a } & = & { 3\over 16 \pi^2 } \bigg\ { { 1613\over 64 } - 24\,\zeta_3 + { 9\over
4}\,\zeta_4 + 15\,\zeta_5 + \left ( - { 691\over 32 } + { 9\over 2}\,\zeta_3\right ) { l_{q\mu}}+ { 105\over 16}\,{l_{q\mu}}^2 - { 3\over 4}\,{l_{q\mu}}^3 \nonumber\\&&\mbox { } + { \bar{m}^2\over q^2}\ , \bigg [ { 1697\over 48 } + 54\,\zeta_3 - 6\,\zeta_4 - 70\,\zeta_5 + \left ( { 13\over 8 } - 24\,\zeta_3\right ) { l_{q\mu}}- { 27\over 4}\,{l_{q\mu}}^2 + 6\,{l_{q\mu}}^3 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{2 } \ , \bigg [ - { 885\over 32 } + 104\,\zeta_3 - 24\,\zeta_4 + 15\,\zeta_5 + 4\,b_4 + \left ( { 1253\over 16 } - 33\,\zeta_3\right ) { l_{qm}}\nonumber\\&&\mbox{\hspace{0.4 cm } } + { 129\over 4}\,{l_{qm}}^2 + 6\,{l_{qm}}^3 + \left ( { 195\over 8 } - { 423\over 4}\,{l_{qm}}- 27\,{l_{qm}}^2\right ) { l_{q\mu}}+ \left ( - { 189\over 4 } + { 81\over 2}\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ - { 421657\over 1944 } + 135\,\zeta_3 - 48\,\zeta_4 + 80\,\zeta_5 + 8\,b_4 + \left ( { 4481\over 27 } + 52\,\zeta_3\right ) { l_{qm}}\nonumber\\&&\mbox{\hspace{0.4 cm } } + { 839\over 6}\,{l_{qm}}^2 + { 1223\over 27}\,{l_{qm}}^3 + \left ( { 909\over 2 } - { 1028\over 3}\,{l_{qm}}- 164\,{l_{qm}}^2\right ) { l_{q\mu}}\nonumber\\&&\mbox{\hspace{0.4 cm } } + \left ( - 192 + 144\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{4 } \ , \bigg [ - { 159609\over 128 } - { 50\over 3}\,\zeta_3 - 120\,\zeta_4 + 280\,\zeta_5 + 20\,b_4 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left ( { 751555\over 1296 } + 203\,\zeta_3\right ) { l_{qm}}+ { 534067\over 864}\,{l_{qm}}^2 + { 38753\over 216}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left ( { 70621\over 32 } - { 3124\over 3}\,{l_{qm}}- { 2695\over 4}\,{l_{qm}}^2\right ) { l_{q\mu}}+ \left ( - { 6225\over 8 } + { 1125\over 2}\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \bigg\ } + \ldots\,\ , , \nonumber \\
\label{eqpi2ap } \\[.4 cm ] \bar{\pi}^{(2),p}_{\it na } & = & { 3\over 16 \pi^2 } \bigg\ { { 14419\over 288 } - { 75\over 4}\,\zeta_3 - { 9\over 8}\,\zeta_4 - { 5\over 2}\,\zeta_5 + \left ( - { 893\over 32 } + { 31\over 4}\,\zeta_3\right ) { l_{q\mu}}+ { 71\over 12}\,{l_{q\mu}}^2 - { 11\over 24}\,{l_{q\mu}}^3 \nonumber\\&&\mbox { } + { \bar{m}^2\over q^2}\ , \bigg [ - { 25\over 144 } + { 44\over 3}\,\zeta_3 + 3\,\zeta_4 + 5\,\zeta_5 + \left ( { 153\over 8 } - 17\,\zeta_3\right ) { l_{q\mu}}\nonumber\\&&\mbox{\hspace{0.4 cm } } - { 119\over 12}\,{l_{q\mu}}^2 + { 11\over 6}\,{l_{q\mu}}^3 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{2 } \
, \bigg [ - { 1417\over 96 } + { 49\over 4}\,\zeta_3 + 12\,\zeta_4 + { 5\over 2}\,\zeta_5 - 2\,b_4 + \left ( { 3925\over 48 } + { 15\over 2}\,\zeta_3\right ) { l_{qm}}\nonumber\\&&\mbox{\hspace{0.4 cm } } + { 38\over 3}\,{l_{qm}}^2 + { 11\over 6}\,{l_{qm}}^3 + \left ( { 397\over 24 } - 49\,{l_{qm}}- { 11\over 2}\,{l_{qm}}^2\right ) { l_{q\mu}}+ \left ( - { 55\over 8 } + { 33\over 4}\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ - { 134501\over 972 } - { 63\over 2}\,\zeta_3 + 24\,\zeta_4 - 20\,\zeta_5 - 4\,b_4 + \left ( { 170609\over 648 } - 40\,\zeta_3\right ) { l_{qm}}\nonumber\\&&\mbox{\hspace{0.4 cm } } + { 497\over 6}\,{l_{qm}}^2 + { 505\over
54}\,{l_{qm}}^3 + \left ( { 13231\over 108 } - { 3473\over 27}\,{l_{qm}}- { 451\over 18}\,{l_{qm}}^2\right ) { l_{q\mu}}\nonumber\\&&\mbox{\hspace{0.4 cm } } + \left ( - { 143\over 6 } + 22\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{4 } \ , \bigg [ - { 3008885\over 3888 } + { 823\over 24}\,\zeta_3 + 60\,\zeta_4 - 70\,\zeta_5 - 10\,b_4 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left ( { 1489679\over 2592 } - 145\,\zeta_3\right ) { l_{qm}}+ { 74807\over 288}\,{l_{qm}}^2 + { 8941\over 216}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left ( { 820055\over 1728 } - { 77761\over 216}\,{l_{qm}}- { 5929\over 72}\,{l_{qm}}^2\right ) { l_{q\mu}}+ \left ( - { 3905\over 48 } + { 275\over 4}\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \bigg\ } + \ldots\,\ , , \nonumber\\ \label{eqpi2nap } \\[.4 cm ] \bar{\pi}^{(2),p}_{\it
l } & = & { 3\over 16 \pi^2 } \bigg\ { - { 511\over 36 } + 4\,\zeta_3 + \left ( { 65\over 8 } - 2\,\zeta_3\right ) { l_{q\mu}}- { 11\over 6}\,{l_{q\mu}}^2 + { 1\over 6}\,{l_{q\mu}}^3 \nonumber\\&&\mbox { } + { \bar{m}^2\over q^2}\ , \bigg [ - { 31\over 36 } - { 4\over 3}\,\zeta_3 + \left ( - { 9\over 2 } + 4\,\zeta_3\right ) { l_{q\mu}}+ { 7\over 3}\,{l_{q\mu}}^2 - { 2\over 3}\,{l_{q\mu}}^3 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{2 } \ , \bigg [ { 53\over 24 } - 12\,\zeta_3 - { 257\over 12}\,{l_{qm}}- { 10\over 3}\,{l_{qm}}^2 - { 2\over 3}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left ( - { 17\over 6 } + 14\,{l_{qm}}+ 2\,{l_{qm}}^2\right ) { l_{q\mu}}+ \left ( { 5\over 2 } - 3\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ { 12551\over 486 } - { 16\over 3}\,\zeta_3 - { 4970\over 81}\,{l_{qm}}- { 133\over
9}\,{l_{qm}}^2 - { 46\over 27}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left ( - { 905\over 27 } + { 988\over 27}\,{l_{qm}}+ { 82\over 9}\,{l_{qm}}^2\right ) { l_{q\mu}}+ \left ( { 26\over 3 } - 8\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{4 } \ , \bigg [ { 573913\over 3888 } - { 4\over 3}\,\zeta_3 - {
13547\over 81}\,{l_{qm}}- { 1315\over 24}\,{l_{qm}}^2 - { 287\over 54}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left ( - { 58285\over 432 } + { 5351\over 54}\,{l_{qm}}+ { 539\over 18}\,{l_{qm}}^2\right ) { l_{q\mu}}+ \left ( { 355\over 12 } - 25\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \bigg\ } + \ldots\,\ , , \label{eqpi2lp } \\[.4 cm ] \bar{\pi}^{(2),p}_{\it f } & = & { 3\over 16 \pi^2 } \bigg\ { - { 511\over 36 } + 4\,\zeta_3 + \left ( { 65\over 8 } - 2\,\zeta_3\right ) { l_{q\mu}}- { 11\over 6}\,{l_{q\mu}}^2 + { 1\over 6}\,{l_{q\mu}}^3 \nonumber\\&&\mbox { } + { \bar{m}^2\over q^2}\ , \bigg [ { 161\over 36 } - { 4\over 3}\,\zeta_3 + \left ( - { 33\over 2 } + 4\,\zeta_3\right ) { l_{q\mu}}+ { 7\over 3}\,{l_{q\mu}}^2 - { 2\over 3}\,{l_{q\mu}}^3 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{2 } \ , \bigg [ { 701\over 24 } + 6\,\zeta_3 - { 293\over 12}\,{l_{qm}}- { 11\over 6}\,{l_{qm}}^2 - { 2\over 3}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left ( - { 17\over 6 } + 14\,{l_{qm}}+ 2\,{l_{qm}}^2\right ) { l_{q\mu}}+ \left ( { 5\over 2 } - 3\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ { 43319\over 486 } + { 596\over 9}\,\zeta_3 - { 3064\over 27}\,{l_{qm}}- { 623\over 27}\,{l_{qm}}^2 - { 92\over 27}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left ( - { 905\over 27 } + { 988\over 27}\,{l_{qm}}+ { 82\over 9}\,{l_{qm}}^2\right ) { l_{q\mu}}+ \left ( { 26\over 3 } - 8\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{4 } \ ,
\bigg [ { 3892289\over 7776 } + { 2333\over 9}\,\zeta_3 - { 20533\over 72}\,{l_{qm}}- { 24871\over 216}\,{l_{qm}}^2 - { 545\over 27}\,{l_{qm}}^3 \nonumber\\&&\mbox{\hspace{0.4 cm } } + \left ( - { 58285\over 432 } + { 5351\over 54}\,{l_{qm}}+ { 539\over 18}\,{l_{qm}}^2\right ) { l_{q\mu}}+ \left ( { 355\over 12 } - 25\,{l_{qm}}\right ) { l_{q\mu}}^2 \bigg ] \bigg\ } + \ldots\,\ , , \label{eqpi2fp } \\[.4 cm ] \bar{\pi}^{(2),p}_{\it s } & = & { 3\over 16 \pi^2 } \bigg\ { { \bar{m}^2\over q^2}\ , \bigg [ - 16\,\zeta_3 - 20\,\zeta_5 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{2 } \ , \bigg [ - 44 + 24\,\zeta_3 + \left ( - 12 - 72\,\zeta_3\right ) { l_{qm}}\bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{3 } \ , \bigg [ { 221\over 8 } + 114\,\zeta_3 + \left ( - { 363\over 4 } - 36\,\zeta_3\right ) { l_{qm}}- 63\,{l_{qm}}^2 - 12\,{l_{qm}}^3 \bigg ] \nonumber\\&&\mbox { } + \left({\bar{m}^2\over q^2}\right)^{4 } \ , \bigg [ { 68146\over 243 } + { 1288\over 9}\,\zeta_3 + \left ( { 7727\over 81 } - 80\,\zeta_3\right ) { l_{qm}}- 86\,{l_{qm}}^2 - 44\,{l_{qm}}^3 \bigg ] \bigg\ } + \ldots\,\ , , \nonumber\\ \label{eqpi2sp}\end{aligned}\ ] ] the results for the one- and two - loop case have been checked against the analytical exact expressions for the polarization functions @xcite in the limit of large external momentum @xmath59 .
as an application of the results of the previous section we consider the decay of a scalar and pseudo - scalar higgs boson into a top - quark pair .
thereby we neglect all light quark masses and denote the @xmath22 top quark mass by @xmath60 and the on - shell one by @xmath13 .
let us in a first step review the born result and the first order qcd corrections .
taking the imaginary part of eqs .
( [ eqpi0s ] ) and ( [ eqpi1s ] ) and transforming the combination @xmath61 into the on - shell scheme @xcite one arrives at ( @xmath62 ) @xmath63 r^{(1),s } & = & 3\ , \bigg\ { { 9\over 4 } + { 3\over 2}\,{l_{ms}}+ { m_t^2\over s}\ , \bigg [ - 6 - 18\,{l_{ms}}\bigg ] + \left({m_t^2\over s}\right)^{2 } \ , \bigg [ - 33 + 36\,{l_{ms}}\bigg ] \nonumber\\&&\mbox { } + \left({m_t^2\over s}\right)^{3 } \
, \bigg [ { 280\over 9 } + { 46\over 3}\,{l_{ms}}\bigg ] + \left({m_t^2\over s}\right)^{4 } \ , \bigg [ { 146\over 3 } + { 45\over 2}\,{l_{ms}}\bigg ] \bigg\ } + \ldots\,\,,\end{aligned}\ ] ] for the scalar case where
@xmath64 is chosen .
for the pseudo - scalar current we get from eqs .
( [ eqpi0p ] ) and ( [ eqpi1p ] ) : @xmath65 r^{(1),p } & = & 3 \ , \bigg\ { { 9\over 4 } + { 3\over 2}\,{l_{ms}}+ { m_t^2\over s } \ , \bigg [ 6 - 6\,{l_{ms}}\bigg ] + \left({m_t^2\over s}\right)^{2 } \ , \bigg [ - 9 - 12\,{l_{ms}}\bigg ] \nonumber\\&&\mbox { } + \left({m_t^2\over s}\right)^{3 } \ ,
\bigg [ - { 260\over 9 } - { 62\over 3}\,{l_{ms}}\bigg ] + \left({m_t^2\over s}\right)^{4 } \ , \bigg [ - { 674\over 9 } - { 361\over 6}\,{l_{ms}}\bigg ] \bigg\ } + \ldots\,.\end{aligned}\ ] ] in fig .
[ figsp01 ] the results are plotted together with the exact expressions . for the abscissa
the variable @xmath66 is chosen .
already the approximations including only the quadratic terms provide a good agreement up to @xmath67 .
it is , however , dramatically improved by incorporating higher terms in @xmath68 , leading to an excellent approximation almost up to @xmath69 for @xmath70 and @xmath71 and up to @xmath72 for @xmath73 .
this serves as a strong motivation to evaluate higher order mass terms for the @xmath1 corrections .
it is interesting to compare the results both for the on - shell and the @xmath74-definition of the quark mass .
therefore we set @xmath64 which is the natural choice because it eliminates all logarithms of the constant and quadratic terms in the @xmath22 scheme . starting from the quartic term ,
however , @xmath75 terms remain , resulting from non - trivial operators which are absent before @xcite .
in contrast to the @xmath22 scheme the results in the on - shell scheme develop @xmath75 terms starting from the lowest order .
it is nevertheless instructive to compare the expansions both in @xmath68 and @xmath76 in the two schemes . therefore the renormalization group invariant quantities @xmath77 and
@xmath25 are considered .
we choose @xmath78 , @xmath79 , @xmath80 gev and @xmath81 gev which corresponds to @xmath82 .
the corresponding @xmath22 top mass , @xmath83 , evaluates to @xmath84 gev and the two - loop beta function leads to @xmath85 . in tab .
[ tabsp ] the numbers are listed .
it can be seen that the quartic terms are still sizeable and in many cases numerically of the same order as the constant and quadratic ones .
they become particularly important when the first two terms happen to cancel to a large extent .
the higher order terms give only small contributions at least in the three - loop case .
usually in the @xmath22 scheme a faster convergence of the qcd series is expected in the sense that the coefficients in front of @xmath86 are smaller than in the on - shell scheme .
considering the sum of all approximation terms available this happens to be true only for the scalar correlator whereas in the pseudo - scalar case it is vice versa .
a possible explanation might be that the coefficient of the leading term in the on - shell scheme is already rather small because of an accidental cancellation between the constant and the logarithm .
for the scalar correlator there is in addition a large cancellation between the first three terms in the @xmath22 scheme so that the sum turns out to be also rather small .
this is not the case for @xmath87 where the first three terms have the same sign .
however , the scheme dependence is drastically reduced by the inclusion of the @xmath1 terms as can be seen in tab .
[ tabsp ] .
finally we want to present numerical values for the physical decay rate into top quarks . for convenience the square of the on - shell mass , @xmath88 ,
is factored out both in the on - shell and @xmath22 scheme .
using eq .
( [ eqwidth ] ) the expansions of the decay width for the scalar and pseudo - scalar case look like : @xmath89 , \\ \bar{\gamma}(h\to t\bar{t } ) & = & \frac{g_fm_hm_t^2}{4\sqrt{2}\pi } \left [ 0.8954 + 0.0604 - 0.0011 + 0.0014 - 0.0020 \right ] , \\ \gamma(a\to t\bar{t } ) & = & \frac{g_fm_hm_t^2}{4\sqrt{2}\pi } \left [ 1.8982 + 0.2941 + 0.0146 + 0.0035 - 0.0051 \right ] , \\ \bar{\gamma}(a\to t\bar{t } ) & = & \frac{g_fm_hm_t^2}{4\sqrt{2}\pi } \left [ 1.7084 + 0.3503 + 0.0903 + 0.0016 - 0.0038 \right],\end{aligned}\ ] ] where the first and second number correspond to the born and @xmath10 correction .
the @xmath1 terms are separated into the non - singlet ( third number ) and singlet ( fourth number ) contribution .
the last number corresponds to the gluonic cut which has to be subtracted .
the last two numbers cancel each other to a large extent which means that the two - gluon - cut dominates the imaginary part of the double - triangle diagram .
furthermore one observes that the two - loop qcd corrections amount up to @xmath90 . to conclude ,
we have computed analytically the first five terms of the scalar and pseudo - scalar current correlator in the expansion for large external momentum @xmath59 . as an application the decay of a scalar and pseudo - scalar higgs boson into top quarks is considered . in this case
the higher order mass corrections , especially the quartic terms , emerge to be important .
we would like to thank k.g .
chetyrkin and j.h .
khn for fruitful discussions and for carefully reading the manuscript .
m.s . would like to thank p. gambino for useful discussions .
the work of r.h . was supported by the `` landesgraduiertenfrderung '' at the university of karlsruhe .
99 p. janot , in _ proceedings of the ringberg workshop : the higgs puzzle what can we learn from lep 2 , lhc , nlc , and fmc ? _ , tegernsee , germany , 813 december 1996 , edited by b.a .
kniehl ( world scientific , singapore , to appear ) .
chetyrkin , r. harlander , j.h .
khn and m. steinhauser , in _ proceedings of the 5th international workshop on new computing techniques in physics research : software engineering , neural nets , genetic algorithms , expert systems , symbolic algebra , automatic calculations ( aihenp 96 ) _ , lausanne , switzerland ,
26 september 1996 ; report nos .
mpi / pht/96 - 122 , ttp96 - 55 and hep - ph/9611354 .
a.h . hoang and t. teubner ,
_ private communication .
broadhurst and s.c .
generalis , preprint out-4102 - 12 , ( 1984 ) ; + k.g .
chetyrkin and v.p .
spiridonov , _ yad . fiz . _ * 47 * ( 1988 ) 818 ; _ sov .
* 47 * ( 1988 ) 522 . | three - loop corrections to the scalar and pseudo - scalar current correlator are calculated . by applying the large momentum expansion mass terms up
to order @xmath0 are evaluated analytically . as an application @xmath1 corrections to the decay of a scalar and pseudo - scalar higgs boson into top quarks
are considered .
it is shown that for a higgs mass not far above the @xmath2 threshold these higher order mass corrections are necessary to get reliable results .
pacs numbers : 12.38.-t , 14.65.ha , 14.80.bn , 14.80.cp .
16.cm @xmath3_institut fr theoretische teilchenphysik , universitt karlsruhe , + d-76128 karlsruhe , germany . + _
@xmath4_max - planck - institut fr physik , werner - heisenberg - institut , + d-80805 munich , germany + _ |
neutron star binary systems have been recognized as potential central engines of gamma - ray bursts ( grbs ) already two decades ago , they have been mentioned in ( 1986 ) and goodman et al .
( 1987 ) and discussed in more detail by eichler et al .
( 1989 ) ( for a more complete bibliography we refer to existing reviews , e.g. meszaros 2002 or piran 2005 ) . as a variant of the neutron star binary case ( 1991 ) discussed systems containing a neutron star ( ns ) and a stellar mass black hole ( bh ) as a possible grb engine .
these days , such compact binary systems are considered the standard model for the subclass of short gamma - ray bursts , that last typically for about 0.3 s ( kouveliotou et al . 1993 ) .
most recently , on 9 may 2005 the first ever x - ray afterglow for a short ( @xmath3 30 ms ) , hard grb ( grb050509b ) has been detected ( bloom et al . 2005 ) .
its tentative association with a nearby giant elliptical galaxy has been interpreted as an indication for a compact binary merger origin of this burst ( bloom et al .
2005 , lee et al .
while nsbh binaries are usually just considered to be a minor variation on the topic of double neutron star merger ( dns ) , it has been pointed out recently ( rosswog et al .
2004 ) that it is not obvious , that such a coalescence will automatically produce a hot and massive accretion disk around the hole .
therefore , its role for gamma - ray bursts needs further investigations .
+ during the disruption process tidal torques are expected to eject material into highly eccentric , possibly unbound orbits .
this debris is extremely neutron rich , @xmath4 , and therefore ( if ejected at appropriate rates ) holds the promise to be one of the still much debated sources of r - process elements ( lattimer and schramm 1974 and 1976 ) .
+ moreover , nsbh systems are generally considered promising sources for ground - based gravitational wave detectors such as ligo @xcite , geo600 ( luck et al .
2001 ) , virgo ( caron et al . 1997 ) and tama ( tagoshi et al . 2001 ) .
+ it is worth pointing out in this context that there is a controversy about the rates at which nsbh mergers do occur .
bethe and brown ( 1998 ) argued that nsbh should merge about an order of magnitude more frequently than dns , while a recent study by pfahl et al .
( 2005 ) comes to the conclusion that the number of nsbh systems in the galaxy should be below 1 % of the number of double neutron star systems . to date 8 dns
have been observed ( stairs 2004 ) , while not a single nsbh binary has been discovered yet .
+ neutron star black hole merger simulations have been performed by several groups .
most of the simulations use newtonian or pseudo - newtonian gravity .
janka et al .
( 1999 ) used a grid based hydrodynamics code together with a nuclear equation of state ( eos ) and a neutrino leakage scheme to explore the role of these mergers in grb context .
lee ( 2000 , 2001 ) and lee and kluzniak ( 1999a , b ) used smoothed particle hydrodynamics with polytropic equations of state to explore the sensitivity of the results to the adiabatic exponent of the eos .
lee and ramirez - ruiz ( 2002 ) have analyzed the flow pattern within an accretion disk around a bh and , in recent papers ( lee and ramirez - ruiz 2004 , lee et al .
2005 ) , they included detailed microphysics in their simulations .
setiawan et al .
( 2004 ) constructed disks around stellar mass black holes and followed their evolution including explicit viscosity , rosswog et al .
( 2004 ) investigated the dynamics of the accretion process .
it is worth pointing out that both setiawan et al .
( 2004 ) and lee et al .
( 2005 ) implicitly assume that interesting disks form whose properties they subsequently investigate whereas the focus of this investigation is the question whether / what kind of a disk forms during the merger .
+ recently , there has been progress in the relativistic treatment black hole neutron star binaries .
taniguchi et al .
( 2005 ) , for example , have been constructing quasi - equilibrium black hole neutron star binaries in general relativity , and further efforts with approximate relativistic treatments seem to be underway ( rasio et al . 2005 ) . + stellar mass black holes formed in core collapse supernovae are thought to be born with masses ranging from about 3 to 20 ( fryer and kalogera 2001 ) .
numerical simulations have so far only explored black hole masses up to 14 ( janka et al . 1999 and lee and kluzniak 1999a , b and lee 2000 , 2001 used black holes up to 10 , rosswog et al .
2004 explored masses up to 14 ) . in this paper
we will focus on black holes with masses ranging from 14 to 20 .
the 14 case has been explored previously ( rosswog et al . 2004 ) using the same microphysics but a purely newtonian bh - potential and may therefore serve to gauge the effect of the -wiita pseudo - potential . + a simple estimate for the radius where a star around a bh is disrupted is the tidal radius , r@xmath5 .
as it grows slower with the black hole mass ( @xmath6 m@xmath7 ) than the gravitational radius ( @xmath6 m@xmath8 ) , it is expected that higher mass bhs will have more difficulties building up massive disks .
therefore , the lower end of the black hole mass distribution is most promising for the launch of a grb .
unfortunately , it is very difficult to find suitable approximations for these cases , as here the space - time is far from a static schwarzschild / kerr solution and therefore neither the use of pseudo - potentials nor solving the hydrodynamics equations in a fixed background metric are admissible .
for these cases fully dynamical general relativistic calculations are needed .
+ in this paper we will explore the accretion dynamics and the observational signatures of the high mass end of these binary systems .
we use a 3d smoothed particle hydrodynamics code that has been developed to simulate compact objects .
most of the code features have been described elsewhere ( rosswog et al .
2000 , rosswog and davies 2002 , rosswog and liebendrfer 2003 ) and will only be briefly mentioned for the sake of completeness . for the simulations described here
we have changed to an integration scheme with individual time steps and also the treatment of the black hole is different from the previous implementation . + in our implementation we have taken particular care to avoid artifacts from the use of artificial viscosity .
the quantities @xmath9 and @xmath10 that are usually used as fixed parameters in the artificial viscosity tensor ( monaghan and gingold 1983 ) , are made time dependent and evolved by solving an additional differential equation ( morris and monaghan 1997 ) . in the absence of shocks @xmath11 ( we always use @xmath12 ) , to be compared with the standard values of @xmath13 and @xmath14 . in strong shocks our @xmath9
is allowed to rise up to values of 2 in order to avoid post - shock oscillations ( rosswog et al . 2000 ) .
in addition , the balsara prescription ( balsara 1995 ) is used to avoid spurious forces in pure shear flows .
thus , artificial viscosity is essentially absent unless a shock is detected .
these measures have proved very effective , suppressing unwanted effects of artificial viscosity by orders of magnitude ( rosswog et al .
2000 ; rosswog and davies 2002 ) .
+ we use a temperature - dependent nuclear equation of state that is based on the tables provided by shen et al .
( 1998a , 1998b ) and that has been smoothly extended to the low - density regime ( rosswog and davies 2002 ) .
it covers a density range in @xmath15 from 0.5 to 15.4 , temperatures from 0 to 100 mev and electron fractions , @xmath16 , from 0 to 0.5 .
the nucleons are treated in the framework of temperature - dependent relativistic mean field theory . at densities below about 1/3 of nuclear matter density nuclei
will be present in the plasma .
its composition is determined for fixed @xmath17 , @xmath18 and @xmath16 from an equilibrium of nucleons , alpha particles and an average heavy nucleus .
thus , energy release due to nuclear transmutations is accounted for in a simple way .
contributions of photons and electrons / positrons of arbitrary degree of relativity and degeneracy are added to the thermodynamic quantities .
details can be found in rosswog and davies 2002 .
+ nowhere ( apart from the initial neutron stars ) @xmath10-equilibrium is assumed .
local changes in the electron fraction and the thermal energy content of matter due to neutrinos are accounted for with a multi - flavor neutrino treatment ( rosswog and liebendrfer 2003 ) .
+ the newtonian self - gravity of the neutron star fluid is calculated using a binary tree ( e.g. benz et al .
this makes the star less compact and might -at least quantitatively- influence the accretion process .
we use a -wiita potential to approximately take into account the presence of general relativistic effects around the black hole such as the presence of a last stable orbit . clearly , the use of the -wiita potential is not a complete substitute for fully fledged general relativistic hydrodynamic simulations around a schwarzschild black hole .
it has , however , turned out to be astonishingly accurate : test - particle orbits with @xmath19 6 m@xmath8 ( geometrical units with g = c=1 are used throughout the paper ) are unstable and orbits with @xmath19 4 m@xmath8 are unbound , i.e. the radius of marginally stable orbit is located at @xmath20= 6 m@xmath8 , and the radius of the marginally bound orbit is r@xmath21= 4 m@xmath8 .
direct comparisons with general relativistic solutions in a schwarzschild space time show that the pseudo - potential is able to capture the essentials of general relativity and can reproduce accretion disk structures to an accuracy of better than 10 @xmath22 ( see e.g. artemova et al .
the pseudo - potential does , however , not prevent velocities from becoming larger than unity .
abramowicz et al . (
1996 ) have suggested a rescaling of the velocities found using newtonian models together with the -wiita potential ( i.e. in our case the numerical velocities ) : @xmath23 , where @xmath24 .
they found this rescaling to reproduce the exact relativistic velocities to better than 5 % .
this could be used , for example , to calculate doppler shifts , this rescaling is however nowhere used in this paper .
+ the build up of the binary tree from scratch is computationally expensive , therefore we do not remove the particles that have crossed our inner boundary ( located at r@xmath25= 3 m@xmath8 ) at each time step .
they are only removed when a dump is written , i.e. every @xmath26 s which is as short as 1/12 of the neutron star dynamical time scale and therefore does not lead to any artifacts ( this has been confirmed by removing particles every time step ) . to avoid numerical problems with the singularity of the -wiita potential
we have extended it smoothly with a polynomial ( see appendix a and fig . [ denom ] ) .
note that all particles that have ever encountered a deviation from the -wiita potential inside r@xmath25 will be removed at the next dump .
+ we have implemented two different time integration schemes with individual time steps : a second - order runge - kutta - fehlberg method ( fehlberg 1968 ) and a second - order predictor - corrector scheme . both of these schemes have been tested extensively against a runge - kutta - fehlberg scheme with a global time step .
as the results were nearly indistinguishable the simulations were run with the predictor - corrector method as it only requires one force evaluation per time step .
the most time consuming part of these simulations is the inspiral where ( due to the stiff neutron star eos and the resulting flat stellar density profile ) practically all particles have to be evolved on the shortest time step .
the gain in speed for the presented calculations is moderate ( factors of a few ) , but in other test problems the code was faster by more than two orders of magnitude . + we construct our initial neutron stars by solving the lane - emden - equations for our eos with the additional constraint of ( cold ) @xmath10-equilibrium .
this yields stellar radii of about 16 km for a 1.4 neutron star .
all sph - particles have the same mass to avoid numerical noise arising from interactions of particles with unequal masses . prior to setting up the binary system the neutron stars are relaxed carefully by applying an additional , velocity proportional damping forces in the equations of motion so that the sph - particles can settle into their equilibrium positions .
once such an equilibrium is obtained the ns - bh - binary system is set up , either with a tidally locked or a non - spinning neutron star .
the locked systems are constructed accurately using the hydrocode itself ( see rosswog et al .
the non - spinning neutron stars are set on orbit with an angular frequency determined by the force balance between centrifugal and gravitational forces . in all cases
the radial velocity component of a point mass binary of the considered separation is added . + the simulations presented here use up to 3@xmath27 sph particles and are currently the best resolved models of neutron star black hole encounters ( rosswog et al .
( 2004 ) used up to 10@xmath28 particles , lee ( 2001 ) used slightly more than 80000 particles ) .
+ rccccccrccccccccc run & m@xmath8/@xmath29 & ns spin & @xmath30 & @xmath31 & r@xmath32 & r@xmath33 & # part . &
@xmath34 + + i & 14/0.1 & c & 36.1 & 127.5 & 124.1 & 117 & 570587 & 34.6 + ii & 14/0.1 & c & 36.1 & 127.5 & 124.1 & 125 & 2971627 & 40.8 + iii & 16/0.0875 & c & 37.7 & 145.5 & 141.8 & 122 & 570587 & 78.1 + iv & 16/0.0875 & c & 37.7 & 145.5 & 141.8 & 123 & 1005401 & 60.9 + v & 18/0.0778 & c & 39.3 & 162.0 & 159.6 & 123 & 570587 & 50.4 + vi & 20/0.07 & c & 40.7 & 187.5 & 177.3 & 128 & 1503419 & 179.6 + + vii & 14/0.1 & i & 36.1 & 126.2 & 124.1 & 116 & 1497453 & 49.8 + viii & 16/0.0875 &
i & 37.7 & 143.9 & 141.8 & 121 & 1497453 & 39.8 + ix & 18/0.0778 & i & 39.3 & 160.4 & 159.6 & 126 & 1497453 & 205.4 + [ runs ]
rccccccrccccccccc run & m@xmath8/@xmath29 & a@xmath8 & l@xmath35 & m@xmath36 & e@xmath37 + + i & 14/0.1 & 0.196 & @xmath38 & 0.20 & @xmath39 + ii & 14/0.1 & 0.200 & @xmath38 & 0.20 & @xmath40 + iii & 16/0.0875 & 0.197 & @xmath41 & 0.15 & @xmath42 + iv & 16/0.0875 & 0.197 & @xmath43 & 0.15 & @xmath42 + v & 18/0.0778 & 0.201 & @xmath44 & 0.08 & @xmath45 + vi & 20/0.07 & 0.198 & @xmath44 & 0.01 & @xmath46 + +
vii & 14/0.1 & 0.203 & @xmath47 & 0.17 & @xmath48 + viii & 16/0.0875 & 0.206 & @xmath49 & 0.11 & @xmath50 + ix & 18/0.0778 & 0.207 & @xmath44 & 0.04 & @xmath51 + [ results ] the dynamics of the coalescence is illustrated in figures [ run ii ] and [ run vi ] , movies can be found at http://www.faculty.iu-bremen.de/srosswog/movies.html . in all of the cases
the neutron star is completely disrupted after a large portion of its mass has been transferred directly into the hole .
the corresponding peak accretion rates exceed 1000 /s for about 1 ms , after this short episode they drop by at least two orders of magnitude , see panel one in figure [ mdot ] . + it is instructive to compare the 14 case to the corresponding case of our previous study ( rosswog et al .
2004 ) , where we had used a newtonian bh potential .
in the purely newtonian case we found episodic mass transfer with a low - mass , `` mini neutron star '' surviving throughout the whole simulation or about eight close encounters . in the -wiita case about 1.15 ( see panel two in figure [ mdot ] ) are transferred directly into the hole , the rest forms a rapidly expanding tidal tail .
the tidal tail still contains an outward - moving density maximum ( corresponding to the mini neutron star of the newtonian case ) , but its self - gravity is not strong enough to form a spherical object .
the results are well converged , runs i and ii show excellent agreement in the bh masses and peak mass transfer rates .
some minor deviations are visible at low mass transfer rates ( see panel one in fig .
[ mdot ] and the distance , @xmath52 , where numerically resolvable mass transfer sets in ; see column seven in table [ runs ] ) .
+ the case with 16 bhs behaves qualitatively very similar to the 14 bhs : slightly more mass ( about 1.2 ) is transferred into the hole , the disk is slightly less massive , hot and dense than the 14 case . again , the two different resolutions yield nearly identical results .
+ the systems containing bhs of 18 or more ( runs v , vi and ix ) do not form accretion disks at all .
almost the complete neutron star flows via the inner lagrange point directly into the hole , only a small fraction of the star is spun up enough by tidal torques to be dynamically ejected , see last column in table [ results ] . in these cases
the remnant consists of the black hole ( without any accretion disk ) and a rapidly expanding , concentric ( half-)ring of neutron - rich debris material ( 0.08 for the 18 and 0.01 for 20 bh ) , see figure [ run vi ] .
+ the result that it seems to be intrinsically difficult to form promising accretion disks ( at least for the investigated mass ratios ) is consistent with recent estimates of miller ( 2005 ) . a hot and thick accretion disk around a black hole
is believed to be an essential ingredient for a grb . as described in the previous section , and as expected from simple analytical estimates , more promising disks form for larger mass ratios , i.e. smaller bh masses . in the described simulations only systems with bh masses below 18 form an accretion disk at all . in the following the most promising case , run
ii ( m@xmath53 , tidal locking ) , is discussed , the @xmath54 case with a non - rotating neutron star ( run vii ) looks very similar .
+ in those cases where a disk forms , substantial parts of it are inside the innermost stable circular orbit at @xmath55 m@xmath8 , and are therefore `` plunging '' with large radial velocities towards the hole which leads to a substantial consumption of the disks during the simulated time . for illustration ,
[ particles_in_disk ] shows the distribution of the sph - particles at t= 18.396 ms of run ii . shown
are the projections of the particles onto the orbital ( = xy- ) plane ( black ) and overlaid are the projections to the xz - plane of those particles @xmath56 with @xmath57 km ( red ) to show the vertical disk structure .
moreover the locus of the schwarzschild radius and @xmath20 are shown .
while the matter cross section far away form the hole is close to circular , the disk close to the hole is geometrically thin ( @xmath58 ) apart from the region close to the spiral shock where the accretion stream interacts with itself . in this regions
the disk is slightly puffed up . due to the plunge motion only moderate densities ( @xmath59 ) and temperatures are reached ( see fig .
[ rho_t_runii ] ) .
the disk is essentially cold apart form the spiral shock where the temperatures reach @xmath60 mev . + the geometrically thin , relatively cool disks that we find for black hole masses between 14 and 16 is in stark contrast to the thick disks that form in the neutron star merger case ( see figs .
15 and 16 in rosswog and davies ( 2002 ) ) .
the annihilation of neutrino anti - neutrino pairs is one possibility to deposit energy in the baryon free region above the hole . in the cases with black hole masses
@xmath61 the neutrino emission negligible ( @xmath62 erg / s ) .
the most promising cases are , as expected , the ones with the lowest black hole masses .
if disks form at all they are completely transparent to the emerging neutrinos . in lowest mass cases with m@xmath53
( runs i , ii , and vii ) we find peak luminosities of @xmath63 erg / s ( where the index @xmath64 refers to the heavy lepton neutrinos ) , i.e. our most promising cases here yield luminosities that are more than two orders of magnitude smaller than in our neutron star merger calculations where the same microphysics was used .
the average energies of the emitted neutrinos are between 12 and 15 mev .
opposite to the neutron star merger case the neutrino luminosity does not settle into a stationary state , since the disk is being consumed on a time scale of tens of milliseconds . in the tidally locked case
we find a single neutrino pulse of about 20 ms duration , in the case without neutron star spin we find two neutrino pulses separated by about 15 ms , see figure [ lnu_tot ] .
we have performed three - dimensional hydrodynamic simulations of neutron black hole encounters with mass ratios @xmath65 .
we used a state - of - the - art temperature - dependent , nuclear eos , a detailed , multi - species neutrino treatment and the -wiita pseudo - potential .
+ we consider all the approximations made to be valid to a high degree .
if an accretion disk forms at all ( i.e. for the bh masses @xmath66 ) it is of only moderate density ( @xmath67 ) and completely transparent to neutrinos . therefore the neutrino emission results can not be influenced by the flux - limited diffusion treatment .
moreover , the results are numerically converged , different numerical resolutions yield almost identical results .
the bhs are massive enough to dominate the space - time completely and as they are spun up to spin parameters of only 0.2 ( see table [ results ] ) , we consider the use of pw - potentials a very good approximation ( note that for a=0.2 the event horizon moves from 2 to 1.98 and the last stable orbit from 6 to 5.33 gravitational radii ; see fig .
[ event_horizon ] ) .
the ejected mass fraction found in these simulations is generally _ very _ large ( see table [ results ] ) , but comparable to the analytical estimates of lattimer and schramm ( 1974 and 1976 ) , who estimated @xmath68 .
we had performed previously nucleosynthetic calculations for the decompressed ejecta in the neutron star merger case ( freiburghaus et al .
as these calculations did not account for @xmath16-changes due to weak interactions , @xmath16 was treated as a free parameter and was varied in a range that is reasonable for neutron star material ( 0.05 to 0.20 ) . for values between 0.08 and 0.15
we found an almost perfect agreement with the observed , solar system r - process abundances from around barium up to beyond the platinum peak .
the value of @xmath16 has basically two effects : i ) it determines the neutron to seed ratio and thus the maximum nucleon number @xmath69 of the resulting abundance distribution and ii ) the location of the reactions in the n - z - plane , the so - called r - process path , which is , in this case , very close to the neutron drip - line .
this path determines the nuclei that are involved and therefore the nuclear energy release and the corresponding @xmath10-decay time scales .
the initial neutron star material will be decompressed via the expansion and will therefore continuously change its composition .
starting from very large initial nuclei with hundreds of nucleons ( the details of which are determined by the distributions of density , temperature and electron fraction ) these transmutations will release nuclear binding energy until a final , @xmath10-stable composition has been reached .
figure [ ye_dist ] shows the snapshot distribution of @xmath16 within the debris of run ii at t= 18.396 ms , in figure [ ye_binned ] @xmath16 has been binned with the ejecta mass for the nsbh cases with q=0.1 . generally , as the only parts that have ever reached high temperatures disappear very quickly into the hole ( see fig .
[ rho_t_runii ] ) , the @xmath16 of the ejected material reflects closely its initial value inside the neutron star .
the found @xmath16s are generally very close to what was found in the neutron star merger case to give excellent results ( compare freiburghaus et al . 1999 , fig .
4 ; note , however , that due to the somewhat different expansion velocities , the interesting @xmath16 values may be slightly different in the nsbh case ) . to determine the exact elemental yields from these mergers is beyond the scope of the current work and is left to future investigations . + for the neutron star merger case argast et al .
( 2004 ) found that a conflict with the observed element ratios in metal - poor halo stars may arise if neutron star mergers were the _
dominant _ r - process source .
similar questions may be raised for the nsbh case .
one possibility to avoid the problem could be that nsbh coalesce much less frequently than dns .
this would be consistent with the non - observation of any nsbh - system ( currently 8 dns systems are known , see stairs 2004 ) and the result of recent studies ( pfahl et al .
2005 ) that estimate the nsbh number in the galactic disk to be less than 0.1 - 1% of the number of dns .
+ we want to point out , however , that the ejecta found in our simulations exhibit very large radial velocities of @xmath70 . as most neutron star black hole binaries are expected to merge in the outskirts of galaxies ( see e.g. perna and belczynski 2002 ) the chances are large that these high - velocity ejecta do not end up within the host galaxies , but rather enrich the intergalactic medium with very heavy , probably high - mass r - process elements .
this interesting issue definitely requires further consideration in the future . the formation of a massive accretion disk around a black hole is thought to be a vital ingredient for a grb central engine .
none of the systems investigated in this study yields disks that are promising as grb engines . as the tidal radius only grows @xmath6 m@xmath7 while the gravitational radius of the bh scales @xmath6 m@xmath8
, low mass black holes will easier form promising disks . in the simulations presented here
, disks only form for bhs with masses below 18 . as mentioned previously , these disks are essentially cold apart from a spiral shock with temperatures of @xmath71 mev .
contrary to the neutron star merger case , these disks are geometrically _ thin _ , of only moderate density and completely transparent to neutrinos .
the peak densities in the shock are @xmath72 ( cgs ) and lower elsewhere ( see fig .
[ rho_t_runii ] ) .
this is to be compared with peak values of @xmath73 ( cgs ) and @xmath74 mev in the neutron star merger case ( rosswog and davies 2002 ) .
moreover , the disk is drained substantially on the simulation time scale of a few 10 ms .
+ in the most promising cases the peak luminosities are about @xmath75 erg / s ( see table [ results ] , fig .
[ lnu_tot ] ) . this is more than two orders of magnitude below the typical neutron star merger case ( rosswog and liebendrfer , 2003 ) .
the hot debris of a double neutron star coalescence deposits @xmath76 erg / s via @xmath77-annihilation in baryon - devoid regions and drives in this way a relativistic , bipolar outflow ( rosswog and ramirez - ruiz 2002 , aloy et al .
the surrounding baryonic material collimates these jets into a small fraction of the solid angle so that they appear as isoptropized @xmath78 erg ( rosswog and ramirez - ruiz 2003 ) . as the neutrino annihilation rate scales roughly with the neutrino luminosity squared
, the most promising cases of this study ( q= 0.1 ) will provide @xmath79 erg / s for about 10 ms .
this estimate is rather on the optimistic side as , due to the thin disk geometry in the nsbh case , the probability for neutrinos to collide close to head - on ( where the annihilation cross - section is maximal ) is severely reduced .
moreover , such disks will not provide powerful winds from neutrino ablation to collimate a possible outflow .
therefore , the signal resulting from neutrino annihilation may be even less luminous .
general relativistic effects like the bending of neutrino trajectories and the redshift of the neutrino energy may influence the result , but their influence goes in opposite directions and will change the non - relativistic results by only a factor of about two ( asano and fukuyama 2001 ) . recently , ramirez - ruiz and socrates ( 2005 ) have argued that neutrino spectra substantially different from black body may increase the overall explosion efficiency by more than an order of magnitude . but even if we stretch all of the available parameters to their most optimistic limits , e.g. assuming one order of magnitude more efficiency from neutrino spectra and an ( ad hoc ) extremely small beaming fraction @xmath80 , it is hard to see how the apparent isotropized luminosity should exceed , say , @xmath81 erg / s ( for about 10 ms ) even in the most promising cases of this study . +
another agent that can possibly feed energy into collimated outflow is the magnetic field .
being naturally endowed by strong magnetic fields to start with the shredding of a neutron star could be expected to be a good candidate for magnetic field amplification .
this may occur via various pathways , the fastest of which is probably the magneto - rotational instability ( velikhov 1959 , chandrasekhar 1991 , balbus and hawley 1998 and references therein ) .
[ beq ] shows the evolution of the disk - averaged magnetic field , @xmath82 making the extreme assumption that the field can reach equipartition field strength .
here , the index @xmath83 runs over the particles inside 600 km ( see figure [ particles_in_disk ] ) around the hole , @xmath84 are the masses , @xmath85 the densities and @xmath86 the sound velocities of the sph - particles . under this assumption an average field strength slightly in excess of @xmath87 g would result . making the most optimistic assumptions
, we will estimate the properties of the possibly resulting fireball ( see piran 1999 for a fireball review ) .
the fireball energy is @xmath88 .
this yields an initial fireball temperature of @xmath89 . the critical value @xmath90 , where @xmath91 with @xmath92 and @xmath93 energy and mass of the fireball , is given by ( piran 1999 ) , @xmath94 .
given the fact that the resulting disk is very thin and neutrino ablation not important , we consider it plausible that the fireball is in the electron opacity dominated range , @xmath95 .
the fireball will thus become transparent to photons before it reaches the matter dominated stage .
as the lorentz - factor grows @xmath96 and the fireball temperature drops @xmath97 , the observed temperature , @xmath98 , will be that of the initial fireball .
this will yield a black body gamma - ray pulse with a temperature of about 3.3 mev . performing similar estimates for the fireball from
neutrino annihilation yields a thermal pulse of about 2 mev .
swift will have difficulties to detect it , but glast should be able to see a pulse of several tens of milliseconds out to a redshift of @xmath99 . + as mentioned previously , these estimates are based on most optimistic assumptions .
as the material plunges into the hole within about one orbit once it has passed r@xmath100 , any instability will have difficulties to amplify the field by large factors .
the expected thermal precursor pulse may therefore be even weaker than estimated above .
+ in summary , none of the investigated systems seems to be an obvious candidate for the central engine of the observed short - hard grbs .
neutron star black hole mergers may still be able to produce grbs
. obviously , smaller mass black holes are more promising .
technically , however , they are much more challenging as the neutron star has a mass comparable mass to the black hole and space - time is far from the schwarzschild solution .
therefore , neither the use of pseudo - potentials nor simulations on a fixed background space - time are justified , here fully relativistic 3d simulations of the full merger process are required .
+ the spin of the black hole may provide a possibility to launch grbs even from more massive black holes .
if they are spinning very rapidly from the beginning and are spun up further during the merger to values close to the maximum spin parameter of @xmath101 then both the position of the last stable orbit and of the event horizon move to 1 m@xmath8 , see fig .
[ event_horizon ] .
therefore , much higher temperatures and densities can be reached in the inner disk regions .
the fraction of viable grb candidates among the nsbh systems obviously depends on their mass and initial bh spin distribution .
both quantities are not observationally known and their predictions via population synthesis models are plagued by large uncertainties .
it is however clear , that not the whole available parameter space will provide conditions that are favorable to launch grbs .
this fact and the non - observation of even a single neutron star black hole system ( while eight double neutron star systems are observed ) makes it plausible that the observed short - hard grbs are dominated by double neutron star coalescences .
+ in the investigated nsbh mergers we find very large amounts of radioactive material being ejected .
the radioactive decay of this debris material would , like in a type ia supernova , power the post - coalescence lightcurve .
the bulk of ejecta material stems from the initial nucleon fluid of the neutron star and will end up in large , possibly r - process , nuclei . if we assume a typical nuclear binding energy of @xmath3 8 mev per nucleon for the final nuclei , the decompression of this material will release , depending on the mass ratio of the nsbh system , between @xmath102 and @xmath103 erg from radioactive decays .
the details of the resulting lightcurve will depend on the nuclear binding energies and decay time scales and their competition with the expansion of the material . as these nuclei will be extremely neutron rich ( see fig .
[ ye_binned ] ) and far away from the valley of @xmath10-stability no experimental data for their half - lives are available and one would have to resort to theoretical predictions . for a simple estimate of the resulting electromagnetic transient from these radioactive ejecta we use the simple analytical estimate of li and ( 1998 ) who assumed spherical symmetry and a uniform distribution of nuclear lifetimes in logarithmic time intervals . scaling their results with the numbers found in our simulations
we find that the peak luminosity of @xmath104 erg / s should be reached about 3 days after the coalescence ( the banana - like geometry may lead to a peak that occurs somewhat later than in the spherical model of li and ) .
the effective temperature at peak will be @xmath105 k and result in an intensity maximum in the optical / near infrared band of the spectrum . + it is , however , worth pointing out that this result relies on radioactive nuclei still being present when the ejecta become optically thin , which depends on the details of the nuclear reaction path .
it may also be possible that the decays have ceased already by that time and the result is an essentially dark explosion. this interesting topic definitely deserves further examination .
+ in summary , the merger of a neutron star with a black hole of a mass beyond 14 may , rather than producing a gamma - ray burst , yield a supernova - like transient with a thermal precursor pulse peaking in the gamma - ray band .
such a precursor of @xmath2 ms duration should be detectable with glast out to a redshift of @xmath99 .
follow - up observations should then be able to detect a few days later the supernova - like transient .
+ movies of the simulations can be found under + http://www.faculty.iu-bremen.de/srosswog/movies.html .
it is a pleasure to thank marek abramowicz , jim lattimer , william lee , andrew macfadyen , enrico ramirez - ruiz , roland speith and christophe winisdoerffer for useful discussions and the infn in catania and the ias in princeton for their hospitality .
the calculations reported here have been performed on the jump supercomputer of the hchstleistungsrechenzentrum jlich .
* appendix a : -wiita potential with smooth extension * + here we show the form of the potential that is used in the simulations . to numerically avoid the singularity at the schwarzschild - radius we extend the -wiita acceleration denominator , @xmath106 , smoothly down to vanishing radius with a polynomial , @xmath107 , that posesses the following properties : + i ) it matches @xmath108 at a transition radius , @xmath109 , : @xmath110 , + ii ) the derivatives match smoothly : @xmath111 , + iii ) its derivative vanishes in the origin @xmath112 .
+ using @xmath113 and @xmath114 the force denominator is expressed as the form of @xmath116 is shown in fig .
[ denom ] , we always use @xmath117 m@xmath8 . note
that every particle that has ever been inside @xmath109 is removed at the next dump step ( the time between two subsequent dumps is a small fraction ( @xmath118 ) of the neutron star dynamical time ) .
+ 99 abramovici a. , althouse w. e. , drever r. w. p. , gursel y. , kawamura s. , raab f. j. , shoemaker d. , sievers l. , spero r. e. , thorne k. s. , science , 256 , 325 ( 1992 ) abramowicz , m.a . ,
beloborodov , a.m. , chen , x.m . and ihumenshchev , i.v .
, a&a , 313 , 334 ( 1996 ) argast , d. , samland , m. , thielemann , f .- k .
and qian , y .. -z . , , 416 , 997 ( 2004 ) asano , k. and fukuyama , t. , apj , 546 , 1019 ( 2001 ) balbus , s.a . and hawley , j.f . ,
phys . , 70 , 1 ( 1998 ) balsara d. , j. comput .
phys . , 121 , 357 ( 1995 ) benz w. , bowers r. , cameron a. , press w. , 1990 , apj , 348 , 647 bloom , j.s .
et al . , astro - ph/0505480 ( 2005 ) chandrasekhar , s. , proc .
sci usa , 46 , 53 ( 1960 ) fehlberg , e. , nasa technical report tr - r287 ( 1968 ) janka , h .-
t . , eberl , t. , ruffert , m. , fryer , c.l . , apj , 527 , l39 ( 1999 ) lattimer , j. and schramm , d.n . , apj , 192 , l145 ( 1974 ) lattimer , j. and schramm , d.n . ,
apj , 210 , l549 ( 1976 ) lattimer , j. et al . ,
apj , 213 , l225 ( 1977 ) lee , w.h . and
kluzniak , w.l . ,
apj , 526 , 178 ( 1999 ) lee , w.h . and kluzniak , w.l . ,
mnras , 308 , 780 ( 1999 ) lee , w.h . , mnras , 318 , l606 ( 2000 ) lee , w.h . ,
mnras , 328 , 583 ( 2001 ) lee , w.h . , ramirez - ruiz , e. and page , d. , astro - ph/0506121 ( 2005 ) lee , w.h . , ramirez - ruiz , e. and granot , j. , astro - ph/0506104 ( 2005 ) miller , m.c . ,
astro - ph/0505094 ( 2005 ) morris j. , monaghan j. , j. comp .
phys . , 136 , 41 ( 1997 ) novikov , i.d . and frolov , v.p . , physics of black holes , dordrecht : kluwer ( 1989 ) perna , r. and belczynski , k. , apj , 570 , 252 ( 2002 ) pfahl , e. , podsiadlowski , p. and rappaport , s. , astro - ph/0502122 piran , t. , phys . rep . , 314 , 575 ( 1999 ) ramirez - ruiz , e. and socrates , a. , astro - ph/0504/257 ( 2005 ) rasio , f. , faber , j. , kobayashi , s. and laguna , p. , astro - ph/0503007 ( 2005 ) rosswog s. , davies m. b. , thielemann f .- k . , piran t. , a & a , 360 , 171 ( 2000 ) rosswog , s. and davies , m .- b . ,
mnras , 334 , 481 ( 2002 ) rosswog , s. and ramirez - ruiz , e. mnras , 336 , l7 ( 2002 ) rosswog , s. and liebendrfer , m. , mnras , 342 , 673 ( 2003 ) rosswog , s. , ramirez - ruiz , e. and davies , m.b . , mnras 345 , 1077 ( 2003 ) rosswog , s. and ramirez - ruiz , e. , mnras , 343 , l36 ( 2003 ) setiawan , s. , ruffert , m. , janka , h .-
, mnras , 352 , 753 ( 2004 ) shen h. , toki h. , oyamatsu k. , sumiyoshi k. , nuclear physics , a 637 , 435 ( 1998 ) shen h. , toki h. , oyamatsu k. , sumiyoshi k. , prog .
, 100 , 1013 ( 1998 ) stairs , i.h . , science , 304 , 547 ( 2004 ) taniguchi , k. , baumgarte , t.w . ,
faber , j.a . and shapiro , s.l . , astro - ph/0505450v2 ( 2005 ) velikhov , e.p . , sov
jetp . , 36 , 995 ( 1959 ) | i discuss simulations of the coalescence of black hole neutron star binary systems with black hole masses between 14 and 20 .
the calculations use a three - dimensional smoothed particle hydrodynamics code , a temperature - dependent , nuclear equation of state and a multi - flavor neutrino scheme .
general relativistic effects are mimicked using the -wiita pseudo - potential and gravitational radiation reaction forces .
+ opposite to previous , purely newtonian calculations , in none of the explored cases episodic mass transfer occurs .
the neutron star is always completely disrupted after most of its mass has been transferred directly into the hole . for black hole masses between 14 and 16 an accretion disk forms , large parts of it , however , are inside the last stable orbit and therefore falling with large radial velocities into the hole
. these disks are ( opposite to the neutron star merger case ) thin and -apart from a spiral shock- essentially cold . for higher mass black holes ( @xmath0 )
almost the complete neutron star disappears in the hole without forming an accretion disk . in these cases
the surviving material is spun up by tidal torques and ejected as a half - ring of neutron - rich matter .
none of the investigated systems is a promising grb central engine .
we find between 0.01 and 0.2 of the neutron star to be dynamically ejected . like in a type ia supernova ,
the radioactive decay of this material will power a light curve with a peak luminosity of a few times @xmath1 erg / s .
the maximum will be reached about three days after the coalescence and will be mainly visible in the optical / near infrared band .
the coalescence itself may produce a precursor pulse with a thermal spectrum of @xmath2 ms duration . |
there has been considerable interest in probing experimentally a theoretical quark - gluon plasma phase transition by employing ultra - relativistic two heavy nuclei scattering .
it is not presently clear whether or not a deconfined plasma has actually been observed .
however , such scattering experiments have apparently produced nuclear matter at very high energy density wherein the matter acts as a relativistic almost _ ideal _ fluid@xcite ; i.e. a quark - gluon matter fluid with very low viscosity@xcite . in previous work@xcite , we argued from the string fragmentation model , that a high energy density fluid may exhibit a glass - like behavior with meson and baryon strings playing a role analogous to polymers in some condensed matter glasses . at very high temperatures ,
the viscosity of the stringy fluid is actually quite small .
it is this regime ( of an almost ideal fluid ) that is of central importance in ultra - relativistic heavy ion scattering .
our purpose is to study how the viscosity of resulting nuclear matter varies with energy density , pressure and baryon density . in sec.[pf ]
we review the kubo formula for the frequency dependent viscosity of a fluid and derive sum rules which allow for the determination of the zero frequency transverse viscosity in terms of the relaxation time for the non - diagonal elements of the pressure tensor . in sec.[qt ] , we consider the quantized unit of circulation which contributes to turbulent eddy currents in the fluid .
a bound for the reynolds number of quantum turbulence leads to a quantum lower bound for the viscosity of the type that has been previously conjectured@xcite .
we review the low energy well established liquid drop model in sec.[oldm ] and argue that the low energy density nuclear fluid is highly viscous .
we turn to high energy nuclear matter in sec.[gcs ] and in particular discuss the formal conformal symmetry associated with glue .
this system is thought to describe high energy qcd nuclear matter in the limit of low baryon density . in sec.[lund
] we review the notion of the qcd inspired string model and employ in sec.[vstring ] the string fragment ion picture to compute the viscosity of high temperature nuclear matter .
the viscosity reaches small values in the ultra high temperature limit . in the concluding sec.[conc ] , we compare the string fragmentation ( jet ) picture with the quark gluon plasma view of high temperature nuclear matter .
for an isotropic fluid in thermal equilibrium , the off diagonal elements of the pressure tensor ( say @xmath0 ) fluctuations are described by the correlation function , @xmath1 wherein @xmath2 .
the complex frequency dependent fluid shear viscosity @xmath3 is determined by the kubo formalism@xcite @xmath4 with @xmath5 as the mass density of the fluid and @xmath6 as the velocity of a finite high frequency transverse sound , one finds the dispersion relations @xmath7 together with the sum rule @xmath8 in the low frequency limit @xmath9 the relaxation time @xmath10 for the decay of the correlation function @xmath11 uniquely determines the shear viscosity via@xcite @xmath12 to find the viscosity at zero frequency for a fluid of mass density @xmath5 employing the kubo formalism , one must determine the velocity of transverse sound @xmath6 and the relaxation time @xmath10 of the off diagonal elements of the pressure .
the following arguments , first applied by feynman@xcite to superfluid helium , hold true for any fluid whose constituent particles all have mass @xmath13 . for flows with eddy currents on a velocity scale @xmath14 and a length scale @xmath15 , the _ circulation _ of the eddy velocity field is given by @xmath16 from quantum fluid mechanics , the feynman number @xmath17 of quantum vortices in a circulating eddy is given by the bohr quantization of circulation @xmath18 i.e. the feynman number is @xmath19 on the other hand , the reynolds number for a circulating eddy is well known to be @xmath20 yielding a ratio which depends only on intrinsic fluid properties and not on the nature of the eddy flow ; @xmath21 if @xmath22 , with @xmath23 as the number of constituent particles per unit volume , then @xmath24 since the feynman number of quantized vortices in an eddy is larger than the reynolds number of the eddy , we have the quantum viscosity inequality @xmath25 the above lower bound is found to apply to all known chemically pure substances .
for example , for pure water @xmath26 .
the first nuclear fluid model was employed in a non - relativistic context . in the liquid drop model ,
one has a momentum sphere of filled quasi - particle states out to a fermi momentum @xmath27 .
the number of nucleons @xmath28 in a spherical nucleus of radius @xmath29 can be computed by filling up a fermi sphere in momentum space and a sphere in position space @xmath30=\frac{8(k_fr)^3}{9\pi } .
\label{oldm1}\ ] ] the liquid drop nucleus is nearly incompressible so that the radius obeys @xmath31 the fermi velocity for a nearly ideal fermi fluid is given by @xmath32 note that @xmath33 which implies that the nucleons within the droplet are not entirely non - relativistic .
the density of energy states per unit volume at the fermi surface is given by @xmath34 the specific heat per unit volume of a fermi liquid is given by @xmath35 which implies a low temperature entropy per unit volume @xmath36 wherein @xmath37 is the excitation energy per unit volume . for weakly excited states
the entropy per unit volume obeys @xmath38 in terms of the total droplet entropy and energy , the thermodynamic equation of state for excitation energy @xmath39 weakly above the ground state energy @xmath40 is given by @xmath41 the density of excited energy levels for the droplet as a whole then reads @xmath42 , \label{oldm9}\ ] ] which is experimentally reasonably accurate for heavy nuclei ; a logarithmic plot of the density of excited states above the ground state yields a square root behavior@xcite , @xmath43 \approx \sqrt{\frac{e - e_{0a}}{\tilde{e}_a } } \ \ { \rm as}\ \ e \to e_{0a}+0^+ . \label{oldm10}\ ] ] in the opposite limit @xmath44 , one finds @xmath45 wherein @xmath46 is the hagedorn temperature@xcite .
the viscosity of the fermi fluid in the liquid drop is quite high .
this may be understood by writing eq.([pf7 ] ) in the form @xmath47 wherein @xmath48 is an appropriate collision time for quasi - particle scattering . from two body scattering phase space considerations ,
one finds that @xmath49 which implies a low nuclear energy liquid drop viscosity given by @xmath50 in which @xmath51 is a dimensionless quantity of order unity , depending on the angular variation of the two body cross section . at very low temperatures ,
the viscosity grows very high .
it will be discussed in what follows that as the temperature grows ever larger , the viscosity diminishes to smaller values until it is quantum limited and the fluid is nearly ideal .
if @xmath52 denotes the gluon field , @xmath53 then the action describing pure glue reads @xmath54 under the conformal scale changes in space time @xmath55 the action is left invariant ; i.e. @xmath56 .
such symmetry dictates that the trace of the energy - pressure tensor @xmath57 is null . in eq.([gcs4 ] ) , @xmath58 represents the energy per unit volume and @xmath59 represents the pressure .
formally , the energy - pressure tensor for glue is given by @xmath60 which implies @xmath61 and which formally verifies the conformal invariance properties @xmath62 on the other hand , for a theory with massive quarks the trace of the energy - pressure tensor is @xmath63 in the presence of gluons , the trace of the energy - pressure tensor can reassert its strength via @xmath64 for example , for a heavy quark , say @xmath65 with mass @xmath13 , in the presence of a low energy gluon field configuration one finds the trace anomaly@xcite @xmath66 where @xmath67 is the qcd strong coupling strength . employing purely thermodynamic reasoning with the gibbs - duhem equation , @xmath68 and the euler relation for the energy density @xmath69 one finds the trace of the energy - pressure tensor @xmath70 if one replaces the quark chemical potentials with the quark activities @xmath71 then the thermodynamic identity for the trace of the energy - pressure tensor has the more simple form @xmath72 or equivalently , the energy density is related to the pressure according to @xmath73 consider the following simple model of a proposed quark - gluon plasma deconfinement .
at very high temperatures , one considers the quarks and gluons to act as a gas exhibiting conformal symmetry .
this implies a pressure @xmath74 and an energy density @xmath75 obeying @xmath76 . \label{gcs16}\ ] ] at low temperatures this ideal behavior can not persist and the conformal symmetry is broken by an energy gap @xmath77 .
the pressure may then be suppressed by a boltzmann factor , @xmath78 which , in virtue of eq.([gcs15 ] ) , implies @xmath79 .
\label{gcs18}\ ] ] plots of the quark - gluon deconfinement based on this model are shown in fig .
[ fig1 ] .
this simple model of the crossover between the confined phase and the quark - gluon plasma phase is in qualitative agreement with lattice qcd computations@xcite .
the theoretical equilibrium statistical thermodynamic arguments presented here are not directly applicable to ultra high energy heavy ion nuclear collision for two reasons : ( i ) very energetic collisions involve non - equilibrium processes and ( ii ) the collision fragments are most often described by qcd string fragmentation models@xcite .
to compute non - equilibrium processes wherein quarks are connected to one another via strings ( gluon electric flux tubes ) we begin with the activation entropy @xmath80 of a single string ; it is@xcite @xmath81+\frac{\tilde{s}}{k_b}\ , \nonumber \\
\frac{\tilde{s}}{k_b } = \frac{1}{2}\left[\ln(3)+ 7\ln\left(\frac{2\pi } { 3}\right)\right ] , \label{lund1}\end{aligned}\ ] ] wherein the hagedorn temperature @xmath46 is related to the string tension @xmath82 via @xmath83 the numerical value of @xmath82 ( and thereby @xmath46 ) is found from the experimental slope of the lowest meson regge trajectory .
the energy of a single string as a function of temperature may be computed via the microcanonical statistical thermodynamic law , @xmath84 yielding , @xmath85 \ \ { \rm where}\ \
\phi=\frac{7k_bt_0}{2}\approx 725\ { \rm mev}. \label{lund4}\ ] ] the plot of energy as a function of temperature is given in fig .
note that the energy always decreases as the temperature increases .
the heat capacity is negative in the physically allowed string temperature range ; i.e. @xmath86 ^ 2<0 , \nonumber \\ 0 >
t > = -\infty\ \ { \rm and}\ \
\infty > t > t_0 .
\label{lund5}\end{aligned}\ ] ] the physical interpretation of eqs.([lund5 ] ) is worthy of note . since in accordance with the second law of thermodynamics ,
stable systems have a positive heat capacity we thereby deduce that a single string _ during a collision _ is in a metastable state .
as discussed in previous work@xcite , the string near the beginning of the collision has a very low energy @xmath87 .
thus , the initial string temperature is negative .
the string _ can not _ have negative energy which implies that string temperatures between zero and the hagedorn temperature are not allowed ; i.e. @xmath88 is unphysical . as the energy of the string energy increases due to the absorption of collision energy @xmath89 , the temperature decreases @xmath90 .
when only a little bit more collision energy is absorbed by the string @xmath91 the string temperature is virtually a positive infinite value .
any further cooling of the temperature requires more and more addition of the collision energy to be absorbed by a single string .
we note in passing , that the notion of a negative temperature going positive by first approaching @xmath92 and reappearing at @xmath93 is _ not _ new .
this property experimentally appears in laboratory nuclear magnetic resonant spin systems and laser media with inverted energy populations . for energy slightly above @xmath94 ,
the string temperature is virtually infinite ; i.e. @xmath95 as @xmath96 . the question then arises ( in the ultra high collision energy per nucleon limit ) as to how much collision energy can be transferred to a single string in order to bring the temperature down from its very high temperature state . as will be made clear in the next sec.[vstring ] , the cooling time exponentially grows larger with decreasing temperature .
furthermore , the higher the collision energy per nucleon , the shorter the duration of a collision .
when the duration of the collision is comparable to the cooling time for a given hot string temperature , the temperature stops getting smaller .
the string temperatures stay high .
while it is certainly possible to create many new strings with the ultra high collision energy , e.g. mesons , it will not be easy to cool the strings down from having a very high temperature .
let us now be more explicit regarding the time - scales in a stringy liquid droplet ; i.e. the _ new _ high energy liquid droplet `` pancake '' model .
the role of the entropy in determining transition rates follows from the the calculation rules of quantum mechanical transition rates ; in particular one averages over initial states and sums over final states . in general
the number of quantum states @xmath97 is related to the entropy @xmath98 by the boltzman formula @xmath99 the ratio of the number of final states to the number of initial states is then determined by the exponential of an entropy difference @xmath100 .
\label{vstring2}\ ] ] transition rates may be phase space dominated by the exponential entropy factors .
the dominant entropy arises from bosonic gluon excitations of the qcd string .
the fermi or bose nature of physical particles depend only on the number , respectively odd or even , of quarks tied together ( in a polymer - like fashion ) by the gluon color electric flux string .
the thermal relaxation times for very hot @xmath95 strings to cool down to finite temperatures strings depends on the exponential of an entropy difference @xmath101 .
\label{vstring3}\ ] ] from eqs.([lund1 ] ) and ( [ lund3 ] ) one finds @xmath102 + \frac{s_{1 \infty}}{k_b}\ , \label{vstring4}\ ] ] wherein @xmath103+\frac{1}{2}\ln 3 .
\label{vstring5}\ ] ] thus the cooling time scale from eqs.([vstring3 ] ) and ( [ vstring4 ] ) may be written as @xmath104 .
\label{vstring6}\ ] ] the viscosity implicit in qcd inspired string models is of a form which follows from eqs.([pf7 ] ) and ( [ vstring6 ] ) ; it is @xmath105 .
\label{vstring7}\ ] ] to compute the attempt frequency @xmath106 for a string to be created , for example between a quark anti quark pair , we need to consider the rate with which a quark in the `` negative vacuum dirac - fermi sea '' is excited into a positive energy state leaving behind a quark hole .
the phase space element along the axis of the string connecting the quark and the hole is @xmath107 wherein the rate of change of momentum of the quark is the force , i.e. the string tension @xmath108 .
thus , the attempt frequency per unit time per unit length to excite a quark anti quark string , i.e. meson , is @xmath109 for string of length @xmath15 , the energy is @xmath110 yielding the attempt frequency @xmath111 wherein eq.([lund4 ] ) has been invoked . from eqs.([vstring7 ] ) and ( [ vstring10 ] ) follows a central result of this work ; i.e. the high energy nuclear fluid viscosity @xmath112 , \label{vstring11}\ ] ] with the ideal quantum viscosity determined as @xmath113 eqs.([vstring11 ] ) and ( [ vstring12 ] ) for the viscosity of a stringy qcd fluid phase is quite similar to the viscosity of a polymer chain glass .
as the temperature is lowered from above to @xmath46 , the viscosity becomes extraordinarily high and the fluid behaves more as a glass then as an ideal fluid . for the ultra high energy collisions of heavy nuclei , the duration of the collision is sufficiently short that @xmath114 .
for such a range of hot string temperatures , the fluid viscosity is extraordinarily low , i.e. the viscosity is the quantum limited value @xmath115 .
the qcd fluid phase is thereby ideal .
we have derived kubo formula sum rules which allowed for the determination of the transverse viscosity in terms of the relaxation time of the pressure tensor .
we then considered the quantized unit of circulation which contributes to turbulent eddy currents in the fluid .
if the the quantum turbulent feynman number @xmath17 is to exceed classical turbulent reynolds number @xmath116 , then there exists a quantum lower bound to viscosity which has been previously conjectured@xcite .
nuclear matter , as it appears in ultra high energy heavy nuclei scattering , was thought by some workers to have the formal conformal symmetry associated with deconfined glue at low baryon density . here ,
the string fragmentation model was employed wherein the quarks and anti quarks are connected to one another by strings whose electric flux tubes consist of glue . the notion of a low baryon density nuclear fluid
is thereby defined as a very stringy gluon liquid with not so many quarks and anti quarks .
this stringy liquid is not quite the same as a deconfined quark gluon plasma .
the viscous properties of a qcd stringy liquid is as described by eqs.([vstring11 ] ) and ( [ vstring12 ] ) .
the stringy liquid is closely analogous to a polymer glass .
while the viscosity grows quite high as the glass ( hagedorn ) temperature is approached from above , for temperatures well above the glass temperature @xmath117 , the viscosity is quite small and quantum limited .
these small viscosity values imply and almost ideal ultra high energy nuclear fluid . | relativistic high energy heavy ion collision cross sections have been interpreted in terms of almost ideal liquid droplets of nuclear matter .
the experimental low viscosity of these nuclear fluids have been of considerable recent quantum chromodynamic interest .
the viscosity is here discussed in terms of the string fragmentation models wherein the temperature dependence of the nuclear fluid viscosity obeys the vogel - fulcher - tammann law . |
the @xcite law , @xmath4 relating the 2d ( projected or surface ) density , @xmath5 , and the dimensionless radius , @xmath6 , is often fit to the luminosity profiles of elliptical galaxies and to the bulges of disk galaxies .
the parameters of the fit include the srsic index , @xmath1 , as well as the constant , @xmath7 , which is normally chosen so that @xmath8 is the radius containing one - half of the projected light ; @xmath9 @xcite . in a recent series of papers , a. graham and co
- workers have shown that the srsic law provides a remarkably good fit to the luminosity profiles of stellar spheroids , from de galaxies to the most luminous ellipticals @xcite .
the fits apply over 2 - 3 decades in radius , and often extend down to the innermost resolvable radius .
deviations from the best - fitting srsic law are typically of order @xmath10 magnitudes rms .
the srsic index @xmath1 is found to correlate well with galaxy absolute magnitude , @xmath11 @xcite , and also with other structural parameters like @xmath8 and @xmath12 @xcite . setting @xmath13 gives the @xcite law , which is a good fit to luminous elliptical galaxies , and @xmath14 is the exponential law , which reproduces well the luminosity profiles of dwarf ellipticals .
there are some known limitations to the applicability of equation ( [ eq : sersic ] ) to the very central regions of some galaxies .
in particular , srsic s law fails to represent adequately the very central profiles of elliptical galaxies with cores ; the pointlike nuclei of some de galaxies ; and the steep power - law density cusps observed in the inner few parsecs of nearby galaxies like m32 and the bulge of the milky way .
the origin of these features is not well understood , but it is likely that they are the result of dynamical processes , possibly involving single or multiple black holes , which act to modify the pre - existing srsic profile in the innermost regions @xcite .
the density profiles of the _ dark matter _
halos formed in @xmath15-body simulations of hierarchical clustering have traditionally been fit to a rather different class of functions , essentially broken power laws @xcite . however , the most recent simulations @xcite suggest that halo density profiles are better represented by a function with a continuously - varying slope . @xcite proposed the fitting function @xmath16 where @xmath17 is the radius at which the logarithmic slope of the _ space _ density is @xmath18 , and @xmath19 is a parameter describing the degree of variation of the slope .
the corresponding density profile is @xmath20 with @xmath21 .
remarkably , this is precisely the same functional form as equation ( [ eq : sersic ] ) with the difference that navarro et al .
fit equation ( [ eq : alpha ] ) to the _ space _ density of dark matter halos , while equation ( [ eq : sersic ] ) applies to the _ projected _ densities of galaxies .
nevertheless the connection is intriguing and a number of questions spring to mind .
does the srsic profile fit the surface density profiles of dark matter halos as well as it fits galaxies ?
we will show here ( 3 ) that the answer is `` yes '' : the same fitting function provides an equally good description of the projected densities of both dark and luminous spheroids . in
4 we ask whether it is most appropriate to fit the srsic law to the space or projected densities of dark matter halos , and whether these functions are better fits than other three - parameter functions .
5 contains some speculations about why a single density law should describe dark and luminous systems over such a wide range in mass .
we constructed nonparametric estimates of the space and projected density profiles of the 19 @xmath0cdm halo models in @xcite , and compared them with a number of fitting functions , including the srsic law , equation ( [ eq : sersic ] ) ; the deprojected srsic law @xmath22 , defined as the spherical density law whose spatial projection is @xmath23 ; and a generalized , three - parameter nfw ( 1996 , 1997 ) profile , which may be expressed as @xmath24 with @xmath25 .
the nfw profile has @xmath26 and @xmath27 ; the @xcite profile has a similar functional form with inner slope @xmath28 .
details of the numerical simulations are given in @xcite .
four halos are `` dwarf '' sized ( @xmath29 ) , seven are `` galaxy '' sized ( @xmath30 ) , and eight are `` cluster '' sized ( @xmath31 ) .
we adopt the notation of that paper ( @xmath32= dwarf , @xmath33=galaxy , @xmath34=cluster ) in what follows .
nonparametric estimates of the space and projected density profiles , @xmath35 and @xmath36 , were constructed using the spherically symmetrized kernels defined by @xcite ( see e.g. @xcite , appendix a ) .
each @xmath15-body point was replaced by a kernel of the form @xmath37 with @xmath38 the width of the kernel associated with the @xmath39th particle and @xmath40 the modified bessel function .
the projected radii @xmath41 were obtained from the @xmath15-body radii @xmath42 by assigning each particle a random position on the sphere of radius @xmath42 .
density estimates were computed on a grid of 100 radial points spaced logarithmically from @xmath43 to @xmath44 ( these radii are defined below ) .
we followed standard practice @xcite and first computed a pilot estimate of the density via a nearest - neighbor scheme , then allowed the @xmath38 to vary as a power @xmath45 of this pilot density . when fitting one of the parametric functions defined above to @xmath46 or @xmath47 , we computed the density estimates on a grid in @xmath48 ( @xmath49 is the geometric mean of the @xmath38 ) to see which choice of kernel parameters minimized the residual of the fit ; typically there was a broad range of @xmath48 values over which the best - fit parameters and their residuals were nearly constant .
the residual was defined as the rms over the radial grid of @xmath50 with @xmath51 the density estimate at grid point @xmath52 and @xmath53 the parametric fitting function ; this is identical to how most observers define the residual .
below we state the rms deviation between the `` measured '' profile and the best - fitting parametric model in terms of magnitudes , denoted by @xmath54 .
we followed the practice in @xcite of only constructing density estimates in the radial range @xmath55 , where @xmath43 is the radius beyond which the halo mass distribution is considered robust to errors or approximations associated with the simulations ( particle softening , relaxation etc . ) and @xmath44 is the virial radius , i.e. the radius within which the mean density contrast is @xmath56 times the critical density .
table 2 of @xcite gives values of @xmath44 and @xmath43 for all halo models .
ccc|cccccc d1 & 3.04 & 0.043 & 3.47 & 0.047 & 5.58 & 0.054 & 1.34 & 0.071 + d2 & 2.63 & 0.043 & 2.89 & 0.024 & 4.47 & 0.029 & 0.89 & 0.088 + d3 & 3.91 & 0.018 & 4.19 & 0.041 & 6.94 & 0.039 & 1.51 & 0.041 + d4 & 2.84 & 0.067 & 3.33 & 0.059 & 5.26 & 0.065 & 1.23 & 0.090 + g1 & 2.94 & 0.030 & 3.17 & 0.036 & 5.38 & 0.038 & 1.23 & 0.047 + g2 & 3.21 & 0.055 & 3.47 & 0.056 & 5.63 & 0.052 & 1.25 & 0.072 + g3 & 2.87 & 0.050 & 3.44 & 0.042 & 5.98 & 0.049 & 1.33 & 0.047 + g4 & 3.30 & 0.040 & 3.70 & 0.022 & 6.13 & 0.015 & 1.36 & 0.027 + g5 & 2.95 & 0.047 & 3.03 & 0.077 & 4.91 & 0.064 & 1.03 & 0.054 + g6 & 2.93 & 0.063 & 3.57 & 0.059 & 6.10 & 0.069 & 1.39 & 0.064 + g7 & 2.82 & 0.051 & 3.09 & 0.087 & 5.04 & 0.097 & 1.22 & 0.110 + c1 & 2.49 & 0.062 & 3.36 & 0.048 & 6.36 & 0.046 & 1.26 & 0.059 + c2 & 2.41 & 0.028 & 2.68 & 0.047 & 4.65 & 0.040 & 1.06 & 0.048 + c3 & 2.50 & 0.031 & 2.92 & 0.026 & 5.02 & 0.031 & 1.18 & 0.029 + c4 & 2.19 & 0.064 & 3.11 & 0.101 & 5.72 & 0.111 & 1.29 & 0.091 + c5 & 2.53 & 0.041 & 2.61 & 0.092 & 4.33 & 0.077 & 0.91 & 0.066 + c6 & 2.16 & 0.044 & 2.62 & 0.050 & 4.49 & 0.065 & 1.10 & 0.064 + c7 & 2.79 & 0.047 & 3.99 & 0.040 & 7.44 & 0.042 & 1.41 & 0.038 + c8 & 1.99 & 0.069 & 2.62 & 0.083 & 4.67 & 0.095 & 1.13 & 0.087 +
with few exceptions , modelling of the luminosity profiles of galaxies is done in projected space .
we therefore began by analyzing the surface density profiles of the dark halos .
nonparametric estimates of @xmath23 for the 19 halos are shown in figure 1a , and figure 1b plots the deviations from the best - fitting srsic model , equation ( [ eq : sersic ] ) ; table 1 gives the best - fitting @xmath1 and @xmath54 .
the mean srsic index is @xmath57 ( @xmath32 ) , @xmath58 ( @xmath33 ) , @xmath59 ( @xmath34 ) , possibly indicating a ( weak ) trend toward decreasing curvature ( lower @xmath1 ) in the profiles of halos of increasing mass .
the @xmath54 values average @xmath60 ( @xmath32 ) , @xmath61 ( @xmath33 ) , @xmath61 ( @xmath34 ) . for comparison ,
@xcite find @xmath62 in a sample of @xmath63 e and s0 galaxies , and @xcite find a mean @xmath54 of @xmath64 in a sample of 12 elliptical galaxies without cores .
the radial range over which luminous galaxies are fit varies from @xmath65 to @xmath66 decades , comparable on average with the @xmath67 decades characterizing our dark - matter halos . while the noise properties are different for the two types of data , the particle numbers in our halo models are small enough ( @xmath68 ) to contribute nonnegligibly to @xmath54 .
we conclude that the srsic law fits dark halos as well as , and possibly even better than , it fits luminous galaxies .
the residuals in figure 1b appear to show some structure .
we will return in a future paper to the question of whether a modification of the srsic law might reduce these residuals even further . :
( b ) deprojected srsic model ; ( c ) equation ( [ eq : alpha ] ) ; ( d ) generalized nfw model , equation ( [ eq : nfw ] ) .
best - fit parameters are given in table 1 . ]
@xcite showed that equation ( [ eq : alpha ] ) provides a good fit to the _ spatial _ density profiles of dark halos .
we showed above ( 3 ) that the srsic law ( [ eq : sersic ] ) is a good fit to the _ surface _ density profiles of dark halos .
an obvious inference is that a _ deprojected _ srsic law should provide a good fit to the space density .
here we ask which function equation ( [ eq : alpha ] ) , or a deprojected srsic law gives a better fit to @xmath69 .
we also consider the quality of fit of another three - parameter function , the generalized nfw profile presented in equation ( [ eq : nfw ] ) . when fitting deprojected srsic profiles to the dark halos , we define @xmath70 to be the srsic index of the projected function ; hence @xmath70 should be close to the index @xmath1 derived when fitting a srsic law to the surface density ( and the two would be equal if the halo s surface density were precisely described by srsic s law ) .
when reporting fits to @xmath69 with equation ( [ eq : alpha ] ) we define @xmath71 , with @xmath19 the shape parameter of equation ( [ eq : drhodr ] ) .
the results are shown in figure 2 and table 1 .
mean values of @xmath54 for the three fitting functions ( deprojected srsic , eq .
( [ eq : alpha ] ) , generalized nfw ) are ( @xmath72 ) for the dwarf halos , ( @xmath73 ) for the galaxy halos , and ( @xmath74 ) for the cluster halos .
thus , the two srsic functions are almost indistinguishable in terms of their goodness of fit : at least over the radial range available , a deprojected srsic profile with index @xmath75 can be well approximated by a srsic profile with @xmath1 in the range @xmath76 .
both functions provide a significantly better fit to @xmath69 than the generalized nfw profile in the case of the dwarf halos , and the two srsic functions perform at least slightly better than nfw for the galaxy halos .
no single function is preferred when fitting @xmath69 for the cluster halos .
another way to compare the halo density profiles with srsic s law is via the radial dependence of the slope .
figure 3 shows nonparametric estimates of the logarithmic slope , @xmath77 , for the dark halos ; slopes were computed via direct differentiation of the kernel density estimates , using a larger kernel width to compensate for the greater noise generated by the differentiation . equation ( [ eq : drhodr ] )
predicts a straight line on this plot . that is a reasonable description of figure 3 .
the value of @xmath77 in the @xmath33 and @xmath34 halos reaches @xmath78 at the innermost radii , consistent with the asymptotic power - law inner behavior of an nfw profile . no obvious convergence to a power law ( constant logarithmic slope ) is seen in figure 3 , and it is likely that simulations of improved resolultion may lead to even shallower slopes at smaller radii , as pointed out by @xcite . from @xcite .
galaxy masses were computed from total luminosities assuming the magorrian et al .
( 1998 ) mass - to - light ratio , with @xmath79 km s@xmath80 mpc@xmath80 . ]
figure 4 shows srsic s @xmath1 ( derived from fits to the surface density ) as a function of mass for our dark halos and for a sample of early - type galaxies .
there is overlap at @xmath29 , the mass characteristic of `` dwarf '' halos and giant ellipticals .
however the galaxies exihibit a much wider range of @xmath1 values , extending to @xmath81 in the case of dwarf ellipticals .
a natural interpretation is that @xmath1 is determined by the degree to which ( dissipationless ) merging has dominated the evolution .
the nearly exponential ( @xmath82 ) profiles of de galaxies are similar to those of disk galaxies , suggesting that dissipation played a critical role in their formation .
luminous ellipticals are the end products of many mergers , the most recent of which are likely to have been gas - poor , and have de vaucouleurs - like profiles ( @xmath83 ) .
this view is supported by numerical simulations ( scannapieco & tissera 2002 ; eliche - moral et al . 2005 ) that show how exponential profiles are converted into de vaucouleurs - like profiles via repeated mergers .
a thornier question is : why should a law like srsic s fit dark or luminous spheroids in the first place ?
srsic s law with @xmath84 has an energy distribution that is roughly boltzmann , @xmath85 , and it is sometimes loosely argued that this `` maximum - entropy '' state is a result of the mixing that accompanies violent relaxation or merging @xcite . with regard to dark halos ,
taylor & navarro ( 2001 ) have shown that the dependence of phase - space density on radius is well approximated by a power law whose corresponding inner density profile has the shallowest slope .
this can again be interpreted as an indication that the halos are well mixed . while our study does not shed a great deal of light on this question
, it does suggest that the scale - free property of srsic s law , @xmath86 , is the feature that links dark and luminous spheroids and that this property may be a hallmark of systems that form via gravitational clustering .
we have shown that the fitting function that best describes luminous galaxies , the srsic law , is an equally good fit to dark halos .
we have not shown that the srsic law is a good fit in an _ absolute _ sense to either sort of system . but given that dark and luminous density profiles are not pure power laws , a three - parameter law like srsic s is as parsimonious a description as one can reasonably expect .
future work should explore whether other , three - parameter fitting functions can describe dark and/or luminous systems better than srsic s law .
we thank alister graham for making available the galaxy data in figure 4 and for informative discussions .
dm was supported by grants ast-0206031 , ast-0420920 and ast-0437519 from the nsf and grant nng04gj48 g from nasa .
jfn acknowledges support from the alexander von humboldt foundation . | we explore similarities in the luminosity distribution of early type galaxies and the mass profiles of @xmath0cdm halos .
the spatial structure of these systems may be accurately described by a simple law where the logarithmic slope of the projected density is a power law of radius ; the srsic law .
we show that this law provides a significantly better fit than a three - parameter generalization of the nfw profile and derive the best - fitting srsic parameters for a set of high - resolution @xmath0cdm halos spanning a wide range in mass .
the mean srsic @xmath1 values are @xmath2 for dwarf- and galaxy - sized halos and @xmath3 for cluster - sized halos , similar to the values that characterize luminous elliptical galaxies .
we discuss possible reasons why the same law should describe dark and luminous systems that span a range of over seven decades in mass . |
challenges posed by the constantly growing urbanisation are complex and difficult to handle .
they range from the increasing dependence on energy , to serious environmental and sustainability issues , and socio - spatial inequalities @xcite .
in particular , we observe the appearance of socially homogeneous zones and dynamical phenomena such as urban decay @xcite and gentrification @xcite that reinforce the heterogeneity of the spatial distribution of social classes in cities . such a segregation characterized by an important social differentiation of the urban space has significant social , economic @xcite and even health costs
@xcite which justify the attention it has attracted in academic studies over the past century . despite the abundant literature in sociology and economics , however , there is no consensus on the adequate way to quantify and describe patterns of segregation . in particular
, the identification of neighbourhoods where the different groups gather is still in its infancy . as stated many times , and at different periods in the sociology literature @xcite
, the study of segregation is cursed by its intuitive appeal . the perceived familiarity with the concept favours what duncan and duncan @xcite called ` naive operationalism ' : the tendency to force a sociological interpretation on measures that are at odds with the conceptual understanding of segregation . as a matter of fact , segregation is a complex notion , and the literature distinguishes several conceptually different dimensions .
massey @xcite first proposed a list of @xmath0 dimensions ( and related existing measures ) , which was recently reduced to @xmath1 by reardon @xcite .
( i ) _ exposure _ which measures the extent to which different populations share the same residential areas ; ( ii ) the _ evenness _ ( and _ clustering _ ) to which extent populations are evenly spread in the metropolitan area ; ( iii ) _ concentration _ to which extent populations concentrate in the areal units they occupy ; and ( iv ) _ centralization _ to which extent populations concentrate in the center of the city .
we identify several problems with this picture .
the first fundamental
issue lies in the lack of a general conceptual framework in which all existing measures can be interpreted .
instead , we have a patchwork of seemingly unrelated measures that are labelled with either of the aforementioned dimensions .
although segregation can indeed manifest itself in different ways , it is relatively straightforward to define what is _ not _ segregation : a spatial distribution of different categories that is undistinguishable from a uniform random situation ( with the same percentages of different categories ) .
therefore , we can define segregation as _ any pattern in the spatial distribution of categories that deviates significantly from a random distribution _
the different dimensions of @xcite then correspond to particular aspects of how a multi - dimensional pattern can deviate from its randomized counterpart .
the measures we propose here are all rooted in this general definition of segregation .
the other issues are technical in nature .
first , several difficulties are tied to the existence of many categories in the underlying data .
historically , measurements of racial segregation were limited to measures between @xmath2 population groups . however , most measures generalise poorly to a situation with many groups , and the others do not necessarily have a clear interpretation @xcite .
worse , in the case of groups based on a continuum ( such as income ) , the thresholds chosen to define classes are usually arbitrary @xcite .
we propose in the following to solve this issue by defining classes in a unambiguous and non - arbitrary way through their pattern of spatial interaction .
applied to the distribution of income categories in us cities , we find @xmath3 emergent categories , which are naturally interpreted as the lower- , middle- and higher - income classes .
second , most authors systematically design a single index of segregation for territories that can be very large , up to thousands of square kilometers @xcite . in order to mitigate segregation , a more local ,
spatial information is however needed : local authorities need to locate where the poorest and richest concentrate if they want to design efficient policies to curb , or compensate for the existing segregation . in other words ,
we need to provide a clear _ spatial _ information on the pattern of segregation .
previous studies @xcite were interested in the characterisation of intra - urban segregation patterns , but they suffer from the limitations of the indicators they use .
in particular , the values they map come with no indication as to when a high value of the index indicates high segregation levels . as a result ,
the maps are not necessarily easy to read .
furthermore , all the descriptions are cartographic in nature and while maps are a powerful way to highlight patterns , we would like to provide further , quantitative , information about the spatial distribution that goes beyond cartographic representation .
the lack of a clear characterization of the spatial distribution of individuals is not tied to the problem of segregation in particular , but pertains to the field of spatial statistics @xcite .
many studies avoided this spatial problem by assuming implicitely that cities are monocentric and circular , and rely on either an arbitrary definition of the city center boundaries , or on indices computed as a function of the distance to the center ( whatever this may be ) .
however , most if not all cities are anisotropic , and the large ones , polycentric ( see @xcite and references therein ) . many empirical studies and models in economics
aim to explain the difference between central cities and suburbs @xcite . yet
, the sole stylized fact upon which they rely city centers tend to be poorer than suburbs ( in the us ) lacks a solid empirical basis .
in the first part of the paper , we define a null model the unsegregated city and define the representation , a measure that identifies significant local departures from this null case .
we further introduce a measure of exposure that allows us to quantify the extent to which the different categories attract or repel one another .
this exposure is the starting point for the non - parametric identification of the different social classes . in the second part
, we define neighbourhoods by clustering adjacent areal units where classes are overrepresented and show that there an increased spatial isolation of classes as population size of cities grows .
we also show that larger cities are richer in the sense that the wealthiest households tend to be overrepresented and the low - income underrepresented in large cities .
finally , we discuss how density is connected to the spatial distribution of income , and how to go beyond the traditional picture of a poor center and rich suburbs .
we focus here on the income distribution , using the data for the 2014 core - based statistical areas
. however , the methods presented in this paper are very general , and can be applied @xcite to different geographical levels , to an arbitrary number of population categories , and to different variables such as ethnicity , education level , etc .
most studies exploring the question of spatial segregation define measures before comparing their value for different cities .
knowing that two quantities are different is however not enough : we also have to know whether this difference is significant . in order to assess the significance of a result , we have to compare it to what is obtained for a reasonable null model .
we assume that we have @xmath4 areal units dividing the city and that individuals can belong to different categories .
the elementary quantity is @xmath5 which represents the number of individuals of category @xmath6 in the unit @xmath7 .
the total number of individuals belonging to a category @xmath6 is @xmath8 and the total number of individuals in the city is given by @xmath9 . in the context of residential segregation
, a natural null model is the _ unsegregated city _
, where all households are distributed at random in the city with the constraints that * the total number @xmath10 of households living in the areal unit @xmath7 is fixed ( from data ) .
* the numbers @xmath8 are given by the data ; the problem of finding the numbers @xmath11 in this unsegregated city is reminiscent of the traditional occupancy problem in combinatorics @xcite .
if we assume that for all categories @xmath6 , we have @xmath12 , they are then distributed according to the multinomial denoted by @xmath13 , and the number of people of category @xmath6 in the areal unit @xmath7 is distributed according to a binomial distribution .
therefore , in an unsegregated city , we have @xmath14 & = n_\alpha\,\frac{n(t)}{n } \\ { \mathrm{var}}\left [ n_\alpha(t ) \right ] & = n_\alpha\,\frac{n(t)}{n } \left ( 1 - \frac{n(t)}{n } \right ) \end{aligned}\ ] ]
the fundamental quantity we will use in the following is the _ representation _ of a category @xmath6 in the areal unit @xmath7 , defined as @xmath15 the representation thus compares the relative population @xmath6 in the areal unit @xmath7 to the value that is expected in an unsegregated city where individuals choose their location at random . or , equivalently , the representation compares the proportion of individuals @xmath6 in the unit @xmath7 to their proportion in the city as a whole . in metropolitan areas
, @xmath8 is large compared to @xmath16 , and the distribution of the @xmath5 can be approximated by a gaussian with the same mean and variance .
therefore we have in the unsegregated case @xmath17 & = 1 \\ { \mathrm{var}}\left [ r_\alpha(t ) \right ] & = \sigma_\alpha(t)^2 = \frac{1}{n_\alpha } \left ( \frac{n}{n(t ) } - 1 \right ) \end{split}\end{aligned}\ ] ] an important merit of the representation is the possibility to define rigorously the notion of _ over-_representation and _ under-_representation of a category @xmath6 in a geographical area . a category @xmath6 is overrepresented ( with a @xmath18 confidence ) in the geographical area @xmath7 if @xmath19 .
a category @xmath6 is underrepresented ( with a @xmath18 confidence ) in the geographical area @xmath7 if @xmath20 .
if the value @xmath21 falls in between the two previous limits , the representation of the category @xmath6 is not statistically different ( at this confidence level ) from what would be obtained if individuals were distributed at random .
existing measures output levels of segregation ( typically a number between 0 and 1 ) but do not indicate whether these levels are _ abnormally _ high . to this respect , the representation is a significant improvement over previous measures . note that the above null model is reminiscent of the ` counterfactuals ' used in the empirical literature on agglomeration economies @xcite .
also , the expression of the representation ( eq . [ eq : repre ] ) is very similar to the formula used in economics to compute comparative advantages @xcite , or to the localisation quotient used in various contexts @xcite . to our knowledge , however , this formula has never been justified by a null model in the context of residential location .
the representation allows to assess the significance of the deviation of population distributions from the unsegregated city . as we will show below , it is also the building block for measuring the level of repulsion or attraction between categories allowing us to uncover the different classes and to identify the neighbourhoods where the different categories concentrate .
last , but not least , the representation defined here does not depend on the category structure at the city scale , but only on the spatial repartition of individuals belonging to each category .
this is essential in order to be able to compare different cities where the group compositions or inequality might differ .
inequality and segregation are indeed two separate concepts , and the way they are measured should be distinct from one another . finally , we would like to mention that using the uniform distribution as a null model can have implications broader than the study of residential segregation . indeed , from a very abstract perspective , the study of residential segregation is the study of labelled objects in space .
the methods presented here can therefore be applied to the study of the distribution of any object in space . in particular , it can be used to identify the locations in a territory where populations with different characteristics ( not necessarily socio - economic ) concentrate .
another shortcoming of the literature about segregation is the lack of indicator to quantify to what extent different populations attract or repel one another .
such a measure of attraction or repulsion is however important to understand the dynamics and scale ( intensity of attraction / repulsion ) of residential segregation .
our indicator is inspired by the m - value first introduced by marcon and puech in the economics literature to measure the concentration of industries @xcite and used as a measure of interaction between retail store categories in @xcite .
these authors were interested in measuring the geographic concentration of different type of industries .
while previous measures ( such as ripley s k - value ) allow to identify departures from a random ( poisson ) distribution , the m - value s interest resides in the possibility to evaluate different industries tendency to co - locate .
the idea , in the context of segregation is simple : we consider two categories @xmath6 and @xmath22 and we would like to measure to which extent they are co - located in the same areal unit . to quantify the tendency of households to co - locate , we measure the representation of the category @xmath22 as witnessed on average by individuals in category @xmath6 , and obtain the following quantity @xmath23 @xmath24 although it is not obvious with this formulation , this measure is symmetric : @xmath25 ( see supplementary information below .
effectively , this ` e - value ' in this context is a measure of exposure , according to the typology of segregation measures proposed in @xcite . however , unlike the other measures of exposure found in the literature @xcite , we are able to distinguish between situations where categories attract ( @xmath26 ) or repel ( @xmath27 ) one another . in the case of an unsegregated city , every household in @xmath6 sees on average @xmath28 and we have @xmath29 .
if populations @xmath6 and @xmath22 attract each other , that is if they tend to be overrepresented in the same areal units , every household @xmath6 sees @xmath30 and we have @xmath31 at the city scale . on the other hand ,
if they repel each other , every household @xmath6 sees @xmath32 and we have @xmath33 at the city scale .
the minimum of the exposure for two classes @xmath6 and @xmath22 is obtained when these two categories are never present together in the same areal unit and then @xmath34 the maximum is obtained when the two classes are alone in the system ( see supplementary information below for more details ) and in this case we get @xmath35 in the case @xmath36 , the previous measure represents the ` isolation ' defined as @xmath37 and measures to which extent individuals from the same category interact which each other . in the unsegregated city , where individuals are indifferent to others when chosing their residence , we have @xmath38 . in contrast , in the extreme situation where individuals belonging to the class @xmath6 live isolated from the others , the isolation reaches its maximum value @xmath39 of course , in order to discuss the significance of the values of exposure and isolation , one needs to compute the variance of the exposure in the unsegregated situation defined earlier .
the calculations for the variance as well as for the extrema are presented in the supplementary information below .
finally , we note that co - location is not necessarily synonymous with interaction , as pointed out by chamboredon @xcite , and we should rigorously talk about _ potential _ interactions .
nevertheless , in the absence of large scale data about direct interactions between individuals , co - location is the best proxy available .
studies that focus on the definition of a single segregation index for cities as a whole can avoid the problem of defining classes , either by measuring the between - neighbourhood variation of the average income ( examples are the standard deviation of incomes @xcite , the variance of logged incomes @xcite and jargowsky s neighbourhood sorting index @xcite ) , or by integrating over the entire income distribution ( for instance the rank - order information theory index defined in @xcite ) .
however , when they investigate the behaviour of households with different income and their spatial distribution , studies of segregation must be rooted in a particular definition of categories ( or classes ) .
unfortunately , there is no consensus in the literature about how to separate households in different classes according to their income , and studies generally rely on more or less arbitrary divisions @xcite .
while in some particular cases grouping the original categories in pre - defined classes is justified , most authors do so for mere convenience reasons . however , as some sociologists have already pointed out @xcite , imposing the existence of absolute , artificial entities is necessarily going to bias our reading of the data .
furthermore , in the absence of recognized standards , different authors will likely have different definitions of classes , making the comparisons between different results in the literature difficult . from a theoretical point of view
, entities such as social classes do not have an existence of their own .
grouping the individuals into arbitrary classes when studying segregation is thus a logical fallacy : it amounts to imposing a class structure on the society before assessing the existence of this structure ( which manifests itself by the differentiated spatial repartition of individuals with different income ) . here ,
instead of imposing an arbitrary class structure , we let the class structure emerge from the data themselves .
our starting hypothesis is the following : _ if there is such a thing as a social stratification based on income , it should be reflected in the households behaviours _ :
households belonging to the same class should tend to live together , while households belonging to different classes should tend to avoid one another .
in other words , we aim to define classes using the way they manifest themselves through the spatial repartition of the different categories .
we choose as a starting point the finest income subdivision given by the us census bureau ( @xmath40 subdivisions ) and compute the @xmath41 matrix of @xmath42 values for all cities .
we then perform a hierarchical clustering on this matrix , succesively aggregating the subdivisions with the highest @xmath42 values .
the process , that we implemented in the python library marble @xcite , goes as follows : 1 .
check whether there exists a pair @xmath6 , @xmath22 such that @xmath43 ( i.e. two categories that attract one another with at least 99% confidence according to the chebyshev inequality ) .
if not , stop the agregation and return the classes ; 2 .
if there is at least one couple satisfying ( 1 ) , normalize all @xmath44 values by their respective maximum values .
find then the pair @xmath45 , @xmath22 whose normalized exposure is the maximum ; 3 . aggregate the two categories @xmath22 and @xmath45 ; 4 . repeat the process until it stops . in order to aggregate the categories at step 3 ,
we need to compute the exposure between @xmath46 and any category @xmath6 , as well as its variance .
the corresponding calculations are presented in the supplementary information ( below ) .
we stress that the obtained classification does not rely on any arbitrary threshold .
indeed , we stop the aggregation process when the only classes left are indifferent ( @xmath29 with @xmath18 confidence ) or repel each other ( @xmath47 with @xmath18 confidence ) . strikingly , the outcome of this method on us data is the emergence of 3 distinct classes ( fig .
[ fig : classes_alluvial ] ) : the higher - income ( @xmath48 of the us population ) and the lower - income classes ( @xmath49 of the us population ) which repel each other strongly while being respectively very coherent and a somewhat small middle - income class ( @xmath50 of the population ) that is relatively indifferent to the other classes .
this result implies that there is some truth in the conventional way of dividing populations into @xmath3 income classes , and that what we casually perceive as the social stratification in our cities actually emerges from the spatial interaction of people .
our method has several advantages over a casual definition : it is not arbitrary in the sense that it does not depend on a tunable parameter ( besides the significance threshold ) and on who performs the analysis .
its origins are tractable , and can be argued on a quantitative basis .
because it is quantitative , it allows comparison of the stratification over different points in time , or between different countries .
it can also be compared to other class divisions that would be obtained using a different medium for interaction , for instance mobile phone communications @xcite . in the following
, we will systematically use the classes obtained with this method . at the scale of an entire country
, segregation can manifest itself in the unequal representation of the different income classes across the urban areas .
we plot on fig .
[ fig : inter - urban_representation ] the ratio @xmath51 where @xmath52 is the number of cities of population greater than @xmath53 , and @xmath54 the number of cities of population greater than @xmath53 for which the class @xmath6 is overrepresented .
a decreasing curve indicates that the class @xmath6 tends to be underrepresented in larger urban areas , while an increasing curve shows that the category @xmath6 tends to be overrepresented in larger urban areas ( the representation is here measured with respect to the total population at the us level ) .
these results challenge sassen s thesis on social polarization , according to which world ( very large ) cities host proportionally more higher - income and lower - income individuals than smaller cities @xcite .
if this thesis were correct , we should observe an overrepresentation of both higher - income and lower - income households in larger cities . instead , as shown on fig .
[ fig : inter - urban_representation ] , higher - income households are overrepresented in larger cities , while lower - income households tend to be underrepresented ( see supplementary information for a detailed discussion ) .
these results support the previous critique of the social polarisation thesis by hamnett @xcite . , scaledwidth=49.0% ]
the representation measure introduced at the beginning of this article allows to draw maps of overrepresentation and thus to identify the areas of the city where categories are overrepresented . in the following ,
we propose to characterise the spatial arrangement of these areas for the different categories . in many studies ,
the question of the spatial pattern of segregation is limited to the study of the center versus suburbs and is usually adressed in two different ways .
in the first case , a central area is defined by arbitrary boundaries and measures are performed at the scale of this central area and the rest is labelled as ` suburbs ' .
the issue with this approach is that the conclusions depend on the chosen boundaries and there is no unique unambiguous definition of the city center : while some consider the central business district @xcite , others choose the urban core ( urbanized area ) where the population density is higher .
the second approach , in an attempt to get rid of arbitrary boundaries , consists in plotting indicators of wealth as a function of distance to the center @xcite .
this approach , inspired by the monocentric and isotropic city of many economic studies such as the von thunen or the alonso - muth - mills model @xcite , has however a serious flaw : cities are not isotropic and are spread unevenly in space , leading to very irregular shapes @xcite . representing any quantity versus the distance to a center
thus amounts to average over very different areas and in polycentric cases ( as it is the case for large cities @xcite ) is necessarily misleading .
as we show below , this method mixes together areas that are otherwise very different .
we propose here a different approach that does not require to draw boundaries between the center and the suburbs .
in fact , it does not even require to define and locate the ` center ' at all .
in the case of a monocentric and isotropic city , our method gives results similar to those given by the other measures . in the more general case where cities are not necessarily monocentric neither isotropic
, our method allows to compare regions of equivalent densities .
the center of a city is usually defined as the region which has the highest population ( or employment ) density .
we therefore propose the density as a proxy to measure of how ` central ' an area is .
we thus plot quantities computed over all areal units ( blockgroups in this dataset ) that have a density population in a given interval @xmath55 $ ] where @xmath56 decreases from its maximum to its minimum value .
we illustrate this idea and compare its results to the traditional ` distance to the center ' method on fig .
[ fig : illustrate_percolate ] . with very anisotropic and polycentric cities such as los angeles ,
the order in which the areal units are considered is very different with both methods . as a result ,
measurements as simple as the average income will yield very different results .
this is particularly striking with the example of seattle , wa shown on fig .
[ fig : illustrate_percolate ] . the average income as a function of the distance to the center ( areal unit with the highest density ) increases from the center to a peak at roughly @xmath57 km ( bottom left figure ) . on the other hand , the bottom left figure shows that average income is , in fact , a simple decreasing function of residential density .
where does the discrepancy come from ?
as one can see on the maps on the top of fig [ fig : illustrate_percolate ] , the units considered in a given distance range can be very different in terms of density . because real cities are neither monocentric or isotropic , units at a same distance from the center can in fact be very different .
this shows , if anything , the importance of expliciting what one means by ` central ' before presenting measures . in the following ,
we express centrality in terms of residential density . here , we compute the representation of groups as a function of residential density .
this method sheds a new light on the difference of social composition between the high - density and low - density areas in cities .
indeed , as shown on fig .
[ fig : high_low_densities ] , we find that low - density regions in cities are on average rich neighbourhoods and that higher density regions are on average lower - income neighbourhoods , in agreement with the dichotomy rich suburbs / poor centers usually found in the literature .
but the dichotomy is not the full picture .
the method indeed entails a surprising result : areas with very large densities ( typically above @xmath58 inhabitant@xmath59 ) are on average rich neighbourhoods .
only few cities in the us have neighbourhoods that reach the threshold of @xmath60 inhabitants per km@xmath61 , which can explain why people have reported in most cases the existence of poor centers and rich suburbs .
in fact , among all @xmath62 blockgroups with a population density greater than @xmath60 inhabitants , @xmath63 are located in new york , ny . most high - density blockgroups belonging to other metropolitan areas exhibit an overrepresentation of the higher - income group and it is thus difficult to conclude at this stage that we are observing an effect specific to manhattan . in any case
, this result suggests that density is very relevant in the usual discussion about income strucutre differences between north american and european cities . of the higher- , middle- and lower - income classes over the @xmath64 cbsa as a function of the local density of households . on average
, we find that low - density regions ( the suburbs ) are on the richer end , while high density regions ( the center ) are on the poorer end .
this confirms on a large dataset a stylized fact that had previously emerged from local studies .
interestingly , we also find that high income households are on average overrepresented in very large density areas ( @xmath65 ) , suggesting that density may be one relevant factor in the explanation of the differences between neighbourhoods .
[ fig : high_low_densities],scaledwidth=49.0% ] intra - unit measures such as the representation or the exposure are not enough to quantify segregation . indeed , areal units where a given class is overrepresented can arrange themselves in different ways , without affecting intra - unit measures of segregation @xcite .
in order to illustrate this , we consider the schematic cases represented on fig .
[ fig : checkerboard ] , and assume that they are obtained by reshuffling the various squares around .
obviously , the checkerboard on the left depicts a very different segregation situation from the divided situation on the right while intra - unit measures would give identical results .
of @xmath66 for the black squares .
( middle ) an intermediate situation between the checkerboard and the divided city , corresponding to @xmath67 .
( right ) the divided city , corresponding to @xmath68 .
[ fig : checkerboard],scaledwidth=49.0% ] defining neighbourhoods in which categories tend to gather is a difficult task .
indeed , what individuals call neighbourhoods often depend on their perception of the city , and field work is often necessary to identify which areas are socially recognised as being a large , middle or low income neighbourhood . however , it is often not possible to do field work and finding a way to define neighbourhoods from population counts is by more convenient and reliable . what is usually defined as a neighbourhood defies the most naive measures .
for instance , to be a member of an ` @xmath6 neighbourhood ' ( where @xmath6 is here higher , middle or low income class ) it is not necessary for an area to have a majority of individuals from the class @xmath6 @xcite .
more sophisticated methods are thus required and the literature is not exempt of such measures , that are all rooted in different assumptions about the nature of neighbourhoods @xcite .
for instance , logan et al .
@xcite use local k - functions in order to assess the prevalence of individuals from a class in an area .
the areas are then clustered using a standard k - means clustering algorithm .
the main issue with this approach is the use of k - functions which measure absolute concentration and are based on the null hypothesis of a completely random distribution of individuals across space . as mentioned earlier in this manuscript , it is more accurate to consider deviations from the null hypothesis of a random distribution of individuals _ among existing locations_. we thus propose here an improvement over logan et al.s definition based on data given at the areal unit level ( but could easily be generalised to data with exact locations ) . as in @xcite , we start with the intuitive idea that an @xmath6 neighbourhood is an area of the city where the category @xmath6 is more present than in the rest of the city . in other words ,
an areal unit @xmath7 belongs to an @xmath6 neighbourhood if and only if the category @xmath6 is overrepresented in @xmath7 , i.e. @xmath69 .
we then build neighbourhoods by aggregating the adjacent areal units where the income class @xmath6 is also overrepresented ( see for example of atlanta fig .
[ fig : atlanta_neighbourhoods ] ) .
a way to distinguish between different spatial arrangements is to measure how clustered the overrepresented areal units are .
the ratio of the number @xmath70 of @xmath6-neighbourhoods ( clusters ) to the total number @xmath71 of areal units when the class @xmath6 is overrepresented ( before constructing the neighbourhood as defined above ) gives a measure of the level of clustering and the quantity @xmath72 is such that @xmath73 in a checkerboard - like situation , and @xmath74 when all areal units form a unique neighbourhood .
we show on fig .
[ fig : clustering ] the distribution of @xmath75 for the three classes over all cities in our dataset .
as one could infer from the maps on fig .
[ fig : atlanta_neighbourhoods ] , the higher - income and lower - income areal units are well clustered , with a respective average clustering of @xmath76 and @xmath77 .
the middle class is on the other hand less spatially coherent , with a average clustering @xmath78 . of the value of the clustering coefficient @xmath79 per class for all cities in our dataset .
the higher - income class exhibits the highest level of clustering , with an average of @xmath80 , followed by the lower - income class with on average @xmath81 .
the middle - income class households are significantly less clustered than the other two , with @xmath82 on average .
the average is computed over all us core - based statistical areas.,title="fig:",scaledwidth=49.0% ] + if a given class is overrepresented in a neighbourhood , it does not necessarily mean that most of the individuals belonging to this neighbourhood are members of this class @xcite .
conversely , the majority of individuals belonging to a class do not necessarily all live in the above - defined neighbourhoods .
we thus compute the ratio of households of each income class that lives in a neighbourhood over the total number of individuals for that income class ( rich , poor , and middle class ) .
our results ( fig .
[ fig : content ] ) indicate that about @xmath83 of the households belonging to @xmath6 live in a @xmath6-neighbourhood , while the rest is dispatched across the rest of the city .
the average concentration decreases from higher - income households ( @xmath84 ) , to lower - income ( @xmath85 ) and middle - income household ( @xmath85 ) .
we compute the fraction of households @xmath86 that belongs to an @xmath6 neighbourhood .
the figure shows the distribution of this fraction for all 2014 cbsas.,title="fig:",scaledwidth=49.0% ] + finally , large values of clustering can hide different situations .
we could have on one hand a ` giant ' neighbourhood and several isolated areal units , which would essentially mean that each class concentrates in a unique neighbourhood . on the other hand
, we could observe several neighbourhoods of similar sizes , meaning that the different classes concentrate in several neighbourhoods across the city . in order to distinguish between the two situations , we plot @xmath87 where @xmath88 is the population of the largest neighbourhood , and
@xmath89 the population of the second largest neighbourhood .
the results are shown on fig .
[ fig : polycentrism ] , and again display a different behaviour for the middle - income on one side , and higher - income and lower - income on the other side .
the size of the middle - income neighbourhoods are more balanced , with on average @xmath90 . in contrast , higher- and lower - income neighbourhoods are dominated by a single large neighbourhood with on average @xmath91 and @xmath92 , respectively .
we compute the ratio of the size of the second largest @xmath6-neighbourhood to the size of the largest @xmath6-neighbourhood .
the above figure shows the distribution of this ratio for all cities in our dataset .
higher- and lower - income households tend to concentrate in a single neighbourhood , with a secondary center that is respectively @xmath93 and @xmath94 the size of the largest one , on average over all cities .
middle - income households tend to be more dispersed , with a secondary neighbourhood that is on average @xmath95 of the size of the largest.,title="fig:",scaledwidth=49.0% ] + as seen in fig .
[ fig : clustering ] , the clustering values are high , indicating that the neighbourhoods occupied by households of different classes are sound . we can now wonder whether there is an effect of the city size on the number of neighbourhoods .
we plot on fig .
[ fig : number_clusters_class ] the number of neighbourhoods found for all three classes as a function of population .
for each class , the curve is well - fitted by a powerlaw function of the form , scaledwidth=49.0% ] @xmath96 where the exponent @xmath97 is less than one and depends on the class , indicating that there are proportionally less neighbourhoods in larger cities ( the number areal units scales proportionally with the population size ) .
in other words , different classes become more spatially coherent as the population increases ( see supplementary information below for more details ) .
the values of the exponents are @xmath98 these values show that the tendency to cluster as the city size increases is larger for higher - income households than for lower- and middle - income households . in other words , segregation increases with city size .
in this paper , we propose a general conceptual framework in which residential segregation can be quantified and understood . instead of enumerating its different aspects
, we took a radically different yet simpler approach .
we define segregation by what it is not : a random distribution of the different households throughout the urban space .
this naturally leads to define the measure of representation , which is used in turn to improve upon previous ways @xcite to define neighbourhoods .
we further define the exposure ( still based on the representation ) , which measures the extent to which different categories attract , repel or are indifferent to one another .
we show that we can define classes in a non - parametric way and 3 main income classes emerge for the 2014 american community survey data .
the middle - income class corresponds to a smaller income range than what is usually admitted , a curiosity that certainly deserves further investigations .
these complex systems can thus be described by considering a small number of categories only .
this is an important piece of information which will simplify the description and modelling of stratification mechanisms . in terms of spatial arrangement , although the fraction of the population that is contained in neighbourhoods does not change with city size , the neighbourhoods are geographically more coherent as cities get larger , which corresponds in effect to an increased level of segregation as the city size increases
. our results also point to the intriguing fact that higher - income households are on average overrepresented in very dense areas .
such high density areas are relatively rare in the us , which might explain in part why we observe poor centers and rich suburbs and rich centers essentially in europe where the density is very large .
this result echoes jacobs analysis @xcite that neighbourhoods with the highest dwelling densities are usually the ones exhibiting the largest vitality , and are therefore the most attractive .
obviously , a high density is not the only determinant and in some cases high - density neighbourhoods can also be lower - income neighbourhoods .
further investigations along these lines may provide quantitative insights into the mechanisms leading to urban decline or urban regeneration . in this study
, we also have tried to highlight the _ spatial _ pattern of segregation .
we believe that the identification of neighbourhoods that our method permits will allow a finer - scale investigation of these spatial patterns . however , the fundamental issue that runs beneath the need for new tools in the analysis of spatial patterns is still open .
it goes beyond the problem of segregation and has a huge number of potential applications .
in this study , we use the american census bureau s 2014 american community survey data on the income of households at the census block - group level , grouped per core - based statistical areas .
the households are divided in 16 income categories , ranging from below @xmath99 annual income to above @xmath100 .
all data of the 2014 american community survey are available from the census bureau .
2140 delineations of the core - based statistical areas are available from the office of budget management .
the reader interested in obtaining a cleaned version of these data ready for analysis and/or reproduce the results of this analysis can consult the online repository ( the code necessary to download , assemble the data and reproduce the analysis performed in this article is freely available online at http://github.com/rlouf/patterns-of-segregation ) .
the methods described in this manuscript are very general , and not limited to the study of income segregation . in order to facilitate their application to other datasets ,
all the measures have been packaged in a python library , marble , open - source and freely available online @xcite .
the following supporting information provide more details on the calculations made to obtain the maximum and minimum value of exposure and isolation , and their variance .
we also detail the process through which we agregate the original categories into classes , and the results we obtain for the @xmath101 american community survey .
finally , we discuss in more details some of the results presented in the manuscript ; in particular , our claim that larger cities are richer .
given two different categories @xmath6 and @xmath22 , we define their intra - unit exposure as the average representation @xmath102 ( resp .
@xmath103 ) of the @xmath6 ( resp .
@xmath22 ) that is seen on average by the members of @xmath22 ( resp .
@xmath6 ) @xmath104 it can also be re - written @xmath105 where @xmath106 this expression symmetric by permutation of the categories @xmath6 and @xmath22 , which is what one would expect from an index measuring the interaction between two categories . in order to know whether the attraction or repulsion mesured between two classes
is significant , we need to be able to compute @xmath107 $ ] . assuming that @xmath108 and @xmath109 are independent ( which is rigorously not true for tracts with a fixed capacity @xmath10 ) , it follows @xmath110 & = { \mathrm{e}}[r_{\alpha}(t ) ] { \mathrm{e}}[r_{\beta}(t)]=1\\ { \mathrm{var } } [ e_{\alpha \beta}(t ) ] & = \frac{1}{n_\alpha n_\beta } \left ( \frac{n}{n(t)}-1 \right)^2 + \frac{1}{n_\alpha } \left ( \frac{n}{n(t ) } - 1 \right ) + \frac{1}{n_\beta } \left ( \frac{n}{n(t ) } - 1 \right)\end{aligned}\ ] ] thus @xmath111 & = 1\\ { \mathrm{var } } [ e_{\alpha \beta } ] & = \frac{1}{n^2 } \sum_t n(t)^2\ , { \mathrm{var}}[e_{\alpha \beta}(t ) ] + \frac{2}{n^2 } \sum_{s < t } n(s)\ , n(t)\ , { \mathrm{cov } } [ e_{\alpha \beta}(s ) , e_{\alpha \beta}(t)]\end{aligned}\ ] ] the covariance is non - zero because the @xmath112 of two different tracts @xmath7 and @xmath113 are not independent , and we have @xmath114 = \left(1-\frac{1}{n_\alpha}\right ) \left(1-\frac{1}{n_\beta}\right ) - 1\ ] ] in order to be able to make sense of the values of exposure ( @xmath44 ) and isolation ( @xmath115 ) , and compare different cities , we need to know their respective maximum and minimum values .
we will consider the following cases : maximum isolation : : situation where each areal unit contains households from one and only one category .
this situation corresponds to the minimum of @xmath44 and the maximum of @xmath115 .
the unsegregated city : : when the distribution of households in the different areal units can not be distinguished from a random distribution .
this is what we call the ` unsegregated city ' and gives a point of reference .
it corresponds to the minimum of @xmath115 . in the unsegregated city case
, there is no way to tell the difference between the distribution of the different categories in the different tracts and a random distribution . in this situation ,
isolation indices @xmath115 reach their minimum value @xmath117 when @xmath118 .
in the maximum isolation case , all categories are alone in their own tract . in other words ,
@xmath119 and @xmath120 we have @xmath121 iff @xmath122 .
we thus obtain for the isolation @xmath123 where @xmath124 is the set of areal units where the category @xmath6 is present . in these unit , @xmath125 .
therefore @xmath126 in the maximum isolation case , all categories are alone in their own tract . in other words ,
@xmath119 and @xmath120 we have @xmath121 iff @xmath122 .
in this situation , we trivially have @xmath127 the maximum of the exposure is however more difficult to obtain in general .
we fix @xmath6 and @xmath22 and we denote by a category @xmath45 all the rest . by definition
we have @xmath128 , @xmath129 , and @xmath130 .
we will look for the ` global ' maximum by keeping the only constraint that in each unit we have @xmath131 .
we obtain for the exposure @xmath132 the maximization of the exposure with respect to @xmath5 thus gives @xmath133\ ] ] which leads to @xmath134 the exposure for these values reads @xmath135 the quantity @xmath136 is in the compact set @xmath137 $ ] and the maximization is not necessarily given by taking the derivative equal to zero .
indeed , in this case the maximum of @xmath138 is obtained for @xmath139 for all @xmath7 ( while the derivative equal to zero would lead to the minimum obtained for @xmath140 for all t ) and reads @xmath141 this maximum is the global one , obtained when there are no constraints .
one can easily add the constraint @xmath128 by using a lagrange multiplier @xmath142 and we have then to maximize the function @xmath143 the derivative with respect to @xmath5 leads to @xmath144 where we expressed the constraint @xmath128 in order to eliminate the lagrange multiplier @xmath142 .
we can then express the maximum @xmath138 obtained for these values of @xmath145 and as above the maximum is obtained for @xmath139 for all t and reads @xmath146\ ] ] which is obviously smaller than the global maximum @xmath147 .
these maxima were obtained when there are no constraints on the total number @xmath148 . when there is such a constraint , the construction of the maximum of eq .
( [ eq : expomax ] ) is not trivial .
very likely , when @xmath149 is fixed , we have to fill the smallest tracts with this class @xmath45 and we are then left with the classes @xmath6 and @xmath22 only .
it seems difficult to obtain an analytical derivation of this maximum and we will keep as a reference in our calculations the global maximum @xmath147 .
the study of segregation must be rooted in a particular definition of class .
however , the income is a continuous variable , and there is no clear definition of incomes classes in the litterature : a class means different things to different people .
we thus start by finding out the class structure as it manifests itself in the spatial arrangement of people .
we take as a starting point the finest income subdivision given by the census bureau ( @xmath40 subdivisions ) and compute the @xmath41 matrix of @xmath44 values at the scale of each cities .
we then perform hierarchical clustering on this matrix , successively aggregating the subdivisions with the highest @xmath44 values .
the process , implemented in the library marble @xcite , goes as follows : 1 .
check whether there exists a pair @xmath6 , @xmath22 such that @xmath43 ( i.e. two categories that attract one another with at least 99% confidence according to the chebyshev inequality ) .
if not , stop the agregation and return the classes ; 2 .
if there are some couples satisfying ( 1 ) , normalize all @xmath44 values by their respective maximum values .
find then the pair @xmath45 , @xmath22 whose normalized exposure is the maximum ; 3 .
aggregate the two categories @xmath22 and @xmath45 ; 4 .
restart the process until it stops . in order to aggregate the categories at step 3 , we need to compute the distance betweence @xmath46 and any category @xmath6 once @xmath22 and @xmath45 have been aggregated . using the definition of @xmath44
, it is easy to show that @xmath150 the variance is also easily calculated as : @xmath151 = \frac{1}{\left ( n_\beta + n_\gamma \right)^2 } \left ( n_\beta^2\,{\mathrm{var}}\left [ e_{\alpha \beta } \right ] + n_\gamma^2\ , { \mathrm{var}}\left [ e_{\alpha \gamma } \right ] \right)\ ] ] we computed the class structure at the scale of the whole us .
we assume that the country is a juxtaposition of the different cities , with independent values of @xmath152 .
we then compute the average over the whole country and obtain @xmath153 where @xmath154 is the population of the city @xmath155 , and @xmath156 the urban population of the us .
the sum runs over all msas in the us . the variance is then given by @xmath157 = \frac{1}{n_{us}^2 } \sum_{c}^{}(n_c)^2\ , { \mathrm{var}}\left [ e_{\alpha \beta}^c \right]\ ] ] starting from categories @xmath158 ( the poorest ) to @xmath159 ( the wealthiest ) , our methods finds the following classes for the us @xmath160 with in parenthesis the percentage of the total us population that is included in the corresponding classes .
although intuitively appealing , the idea that larger metropolitan areas are richer is not as straightforward as it seems . the first question one can ask is if people are richer on average in large cities ? as shown in fig .
[ fig : scaling_income ] , the total income in a city scales ( slightly ) superlineary with population size @xmath161 which suggests that the income _ per household _ is on average higher in larger cities than in smaller ones . in other words , there are proportionally more households belonging the to the wealthiest categories in large cities . in other words ,
the income inequality is higher in large cities than in small ones .
( @xmath162 ) .
[ fig : scaling_income],scaledwidth=50.0% ] in order to measure levels of income inequality , we compute the gini coefficient of the income distribution for every core - based statistical area using the formula proposed in @xcite @xmath163 msa in @xmath101 versus the number of households in the city . no clear trend can be observed here.[fig : gini],scaledwidth=50.0% ] the results , shown in fig .
[ fig : gini ] do not show any dependence of the gini coefficient on the metropolitan population .
this example shows that the gini coefficient is not always a good measure of inequality and can be too aggregated to detect finer details . in order to confirm the consequence of the superlinear scaling of income in terms of larger cities having proportionally more higher - income households
, we plot the number of households belonging to the @xmath3 different classes as a function of the total number of households on fig .
[ fig : scaling_class ] .
we find that for three classes , the data are well approximated by a power - law relationship @xmath164 the problem with writing scaling relationships in this case is that it the constraint @xmath165 is hidden ( ie .
the numbers of households belonging to each category must sum to the total number of households ) .
we therefore write @xmath166 where @xmath167 is the fraction of households in the city that belong to the class @xmath168 .
the constraint that the numbers of households in each class should sum to @xmath53 is equivalent to @xmath169 we plot these ratios on fig .
[ fig : ratio_class ] and we indeed see that the number of households belonging to the higher - income class is proportionally larger in larger cities ( for @xmath170 ) , while the number of households belonging to the lower - income class is proportionally smaller .
the proportion of middle - income class households stays essentially the same across all metropolitan areas . in this work
, we take a different approach and ask if the different classes are more or less represented in a given msa , compared to the average us result . in this context , a city is richer if the higher - income class is over - represented in this city , while the lower - income class is under - represented .
the measure stems from the realisation that rich and poor are not absolute concepts , but must be related to the environment . in this case , it makes sense to compare the representation of the different income classes between metropolitan areas .
neighbourhoods identify the areas in the city where the categories are overrepresented , but this does not necessarily mean that most households belonging to a category live in either of the corresponding neighourhood .
we plot the distribution of the proportion of households belonging to the lower- , middle- and higher- income classes that also live in a corresponding neighbourhood on fig .
[ fig : ratio ] .
one can see that higher - income households tend to be more concentrated in the regions where they are represented , with an average of @xmath84 . followed by the lower - income households , with and average of @xmath85 .
the middle - income households are equally evenly spread across the city , with an average of @xmath85 .
, scaledwidth=50.0% ]
in the main text , we find that the number of neighbourhoods for the @xmath3 classes grows sublinearly with the size of a city , with a behaviour that is well approximated by a power - law with @xmath172 ( @xmath173 ) for all classes together .
we claim this shows the tendency of classes to cluster more in larger cities than in smaller ones .
this is only true , however , if the number of areal units in which each class is overrepresented does not itself vary sublinearly with population size .
we plot on fig .
[ fig : overrepresented ] these numbers for each class and each city as a function of the size of the city .
we find that the behaviour of the number of overrepresented units is consistent with a linear behaviour for all three classes massey , douglas s. , mary j. fischer , william t. dickens , and frank levy .
the geography of inequality in the united states , 1950 - 2000 [ with comments ] .
brookings - wharton papers on urban affairs . 2003;140 . | the spatial distribution of income shapes the structure and organisation of cities and its understanding has broad societal implications . despite an abundant literature ,
many issues remain unclear . in particular , all definitions of segregation are implicitely tied to a single indicator , usually rely on an ambiguous definition of income classes , without any consensus on how to define neighbourhoods and to deal with the polycentric organization of large cities . in this paper , we address all these questions within a unique conceptual framework .
we avoid the challenge of providing a direct definition of segregation and instead start from a definition of what segregation is not .
this naturally leads to the measure of representation that is able to identify locations where categories are over- or underrepresented .
from there , we provide a new measure of exposure that discriminates between situations where categories co - locate or repel one another .
we then use this feature to provide an unambiguous , parameter - free method to find meaningful breaks in the income distribution , thus defining classes .
applied to the 2014 american community survey , we find 3 emerging classes low , middle and higher income out of the original 16 income categories .
the higher - income households are proportionally more present in larger cities , while lower - income households are not , invalidating the idea of an increased social polarisation .
finally , using the density and not the distance to a center which is meaningless in polycentric cities we find that the richer class is overrepresented in high density zones , especially for larger cities .
this suggests that density is a relevant factor for understanding the income structure of cities and might explain some of the differences observed between us and european cities . |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.